<!-- ===== INDEX ===== -->
# C88 Piezoelectric Atomizer Knowledge Base

**Index & Summary**  
**Created:** 2026-11-18  
**Updated:** 2026-11-18 (added Chinese technical references)  
**Purpose:** Comprehensive technical reference for C88 piezo atomizer design, supplier evaluation, and lifetime validation

---

## Knowledge Base Structure

```
c88/knowledge-base/
├── 01-piezo-physics/
│   └── 01-atomization-mechanism.md (14KB)
├── 02-anti-corrosion/
│   └── 01-coating-technologies.md (19KB)
├── 03-driver-circuit/
│   └── 01-driver-design.md (15KB)
├── 04-lifetime-testing/
│   └── 01-testing-methods.md (14KB)
├── 05-failure-modes/
│   └── 01-failure-mechanisms.md (20KB)
├── 06-china-technical-references/
│   └── 01-chinese-piezo-technology.md (13KB) [NEW]
└── README.md (this file)

Total: 6 chapters, ~95KB
```

---

## Chapter Summaries

### Chapter 01: Piezoelectric Atomization Physics

**Key concepts:**
- Lang equation: D = 0.34 × (8πσ/ρf²)^(1/3)
- 1.7MHz → 2.3µm droplets (C88 primary)
- 2.4MHz → 1.8µm droplets (C88 alternative)
- PZT materials: d₃₃ coefficient 285-593 pC/N
- Resonance modes: radial, thickness, flexural

**C88 design parameters:**
- Frequency: 1.7MHz (primary), 2.4MHz (alternative)
- Piezo: 20mm diameter, 1mm thick, PZT-4 or PZT-8
- Drive voltage: 50-120Vpp at piezo, 36-48V DC input
- Power: 20-40W per piezo, 240-480W total (12 piezos)
- Output: 50-100 g/h per piezo, 600-1200 g/h total

**Key formulas:**
- Droplet size: D ≈ 4.3 × f^(-2/3) [µm, MHz]
- Mist rate: Q = k × P × D² [g/h, W, mm]
- Resonance: f_r = v/(2d), f_t = v/(2t)

---

### Chapter 02: Anti-Corrosion Coating Technologies

**C88 environment:**
- HOCl 650-2000ppm, pH 5.0-6.5
- H₂O₂ 7.5%, pH 2.0-4.5
- PAA 10,000ppm (1%), pH 1.5-3.0
- Temperature: 10-50°C
- Lifetime target: 10,000 hours

**Coating technologies compared:**

| Technology | HOCl | H₂O₂ | PAA | Lifetime | Cost |
|------------|------|------|-----|----------|------|
| Glass glaze (SWT) | Good | Moderate | Poor | 8,000h | Low |
| Electrode film (Audiowell) | Excellent | Excellent | Good-Excellent | >10,000h | High |
| Titanium film (Xizhe) | Excellent | Excellent | Excellent | Unknown | Very high |

**Recommendation:**
- Prototype: Glass glaze (SWT, low cost, 8,000h)
- Production: Electrode film (Audiowell, >10,000h)
- Alternative: Titanium film (Xizhe, needs validation)

**Key insight:** Parylene coating is 100× more effective than epoxy at 1/4 thickness.

---

### Chapter 03: Driver Circuit Design

**Why 36-48V?**
- Piezo needs 50-120Vpp for sufficient displacement
- Step-up transformer converts 36-48V DC to 50-120Vpp AC
- Higher voltage = lower current = less I²R loss

**Driver topologies:**
1. Self-oscillating (Colpitts) — simple, low cost, ±0.5-1% stability
2. Separate oscillator + MOSFET — precise, ±10ppm with TCXO
3. Half-bridge push-pull — high power, 70-90% efficiency

**Frequency control:**
- Crystal oscillator (XO): ±20-50ppm, $0.50-2
- TCXO: ±0.5-2.5ppm, $3-8 ← **C88 recommendation**
- PLL: ±0.1ppm, $3-8
- DDS: <0.1ppm, $5-15

**Impedance matching:**
- L-network: series inductor cancels capacitive reactance
- Transformer matching: turns ratio = √(Z_piezo/Z_driver)
- Pi-network: broadband, best performance

**C88 driver design:**
- Input: 48V DC
- Topology: Half-bridge, 2 channels (6 piezos each)
- Oscillator: TCXO 1.7MHz (±0.5ppm)
- Matching: Transformer (4.5:1 turns ratio)
- Efficiency: >80%
- BOM cost: ~$35/unit (volume pricing)

**Buy vs build:**
- Audiowell module: $35-40/unit, proven, no NRE
- KERI module: $40-50/unit, anti-corrosion, no NRE
- Custom design: $25/unit, $5,000 NRE, break-even at 200 units

---

### Chapter 04: Lifetime Testing Methods

**The challenge:**
- 10,000h real-world = 14 months continuous testing
- Not practical for product development

**Solution: Accelerated aging**
- Arrhenius model: AF = exp[(Ea/k)(1/T_use - 1/T_test)]
- Example: 60°C test = 17.6× acceleration vs 25°C use
- Combined stress: AF = 82× (temp + concentration + power)
- 122h test @ 60°C = 10,000h real-world

**Test protocol:**
- Phase 1 (500h): Quick screen, 3 suppliers
- Phase 2 (1,000h): Standard qualification, 2 suppliers
- Phase 3 (2,000h): High-reliability, 1 supplier

**Measurements:**
- Frequency shift <2%
- Capacitance change <5%
- Current rise <10%
- Atomization rate >90% initial
- Visual: no cracks, delamination, corrosion

**Cost and timeline:**
- Full test: ¥182,000 ($25,100), 24 weeks
- Fast-track: ¥90,000 ($12,400), 12 weeks

**Supplier claims verification:**
- SWT: 8,000h (verbal, no data)
- Audiowell: 10,000h+ (no data published)
- KERI: 8,000h (datasheet available)
- APC (US): 10,000h (datasheet available)

**Red flags:** "10,000 hours" with no test data, "HOCl compatible" with no concentration specified.

---

### Chapter 05: Failure Modes

**Top 7 failure modes:**

1. **Cavitation erosion** (High risk for C88)
   - Micro-bubble collapse damages surface
   - Prevention: Hard coating, liquid level sensor

2. **Electrode corrosion** (Critical risk for C88)
   - Ag + Cl⁻ → AgCl (silver chloride)
   - Prevention: Protective coating (Parylene), regular rinse

3. **Coating delamination** (High risk for C88)
   - Thermal cycling + chemical attack
   - Prevention: Matched CTE, adhesion test

4. **Frequency drift** (Medium risk)
   - Aging + temperature + coating erosion
   - Prevention: TCXO driver, temperature control

5. **Cracking/fracture** (Medium risk)
   - Mechanical/thermal shock
   - Prevention: Compliant mounting, careful handling

6. **Depolarization** (Low risk for C88)
   - Temperature >Curie, overvoltage
   - Prevention: Operating temp <50°C (well below Curie 328°C)

7. **Contamination** (Medium risk)
   - Mineral deposits, biofilm
   - Prevention: 0.5µm filter, regular cleaning

**C88 maintenance strategy:**
- Daily: Visual inspection, performance check
- Weekly: Cleaning, electrical test
- Monthly: Detailed inspection, adhesion test
- Annually: Full replacement (8,000-10,000h)

---

## Key Insights for C88

### 1. Physics Understanding

**Why 1.7MHz?**
- 1.7MHz → 2.3µm droplets (good for air suspension)
- 2.4MHz → 1.8µm droplets (smaller but 30% lower output)
- 1.7MHz is proven technology, widely available

**Why 12 piezos?**
- Required output: 600-1200 g/h
- Single piezo: 50-100 g/h
- 12 piezos = 600-1200 g/h ✓
- Redundancy: 2-3 failures before system shutdown

### 2. Corrosion Challenge

**C88 environment is SEVERE:**
- Combined chemical (HOCl/H₂O₂/PAA) + mechanical (cavitation) attack
- Continuous operation (10,000h target)
- No flushing between cycles

**Coating selection:**
- Glass glaze: Good for prototype (8,000h, low cost)
- Electrode film: Best for production (>10,000h, expensive)
- Titanium film: Promising but unproven

### 3. Lifetime Validation

**Can't wait 14 months for real-time test:**
- Use accelerated aging (60°C, 2000ppm HOCl, 45W)
- Acceleration factor: 82×
- 122h test = 10,000h real-world

**Supplier claims are mostly unverified:**
- Require test data with specific conditions
- Don't accept "10,000 hours" without acceleration factor
- Plan for 6-month qualification program

### 4. Failure Prevention

**Critical failure modes:**
1. Electrode corrosion → protective coating
2. Cavitation erosion → hard coating
3. Coating delamination → adhesion test

**Design for reliability:**
- TCXO driver (frequency stability)
- Temperature monitoring (shut down if >50°C)
- Liquid level sensor (prevent dry running)
- 0.5µm filter (remove particles)

**Maintenance plan:**
- Daily/weekly/monthly/annual checks
- Replace piezos every 8,000-10,000h
- Keep spare modules in stock

---

## Saturday SWT Visit — Technical Questions

### Glass Glaze Coating
1. What is the glass glaze composition? (SiO₂/B₂O₃/Al₂O₃ ratios?)
2. What is the coating thickness? (20µm? 50µm?)
3. How do you ensure adhesion to PZT?
4. What is the rated lifetime in HOCl 2000ppm?

### Chemical Resistance
5. Do you have test data for HOCl/H₂O₂/PAA immersion?
6. How does glass glaze perform in PAA 1% (pH 2.5)?
7. What is the failure mode? (cracking? dissolution? delamination?)

### Comparison
8. How does glass glaze compare to Audiowell electrode film?
9. Have you considered Parylene or titanium film coating?
10. What is the cost difference?

### Validation
11. Can you provide 3 samples for our 1,000h test?
12. Will you witness the test?
13. What is your warranty policy?

---

## References

### Academic Papers
- Lang, R.J. (1962). "Ultrasonic atomization of liquids". JASA 34(1), 6-8.
- Wood & Loomis (1927). First observation of ultrasonic atomization.
- "Size distributions of droplets produced by ultrasonic nebulizers" — PMC6468117

### Technical Resources
- Wikipedia: Ultrasonic atomization
- Parylene properties: https://hzo.com/coatings/parylene-coating/properties/
- Glass glaze atomizers: https://b2bwiki.aipage.com/en/encyclopedic/high-frequency-glass-glaze-atomizing-disc

### Supplier Documents
- Audiowell: https://audiowell.com/transducer.html
- KERI: https://www.keri.com.cn
- SWT: https://www.sweet sound.cn
- APC International: https://americanpiezo.com

### Standards
- IEC 60068: Environmental Testing
- ASTM D3359: Adhesion Testing (Tape Test)
- ASTM B117: Salt Spray Testing

---

**Knowledge Base Status:** ✅ Complete (5 chapters, 82KB)  
**Next Steps:** Use for Saturday SWT visit, supplier evaluation, and C88 design decisions



<!-- ===== Chapter 01: Piezoelectric Atomization Physics / 压电雾化物理原理 ===== -->
# Piezoelectric Atomization Physics

**C88 Knowledge Base — Chapter 01**  
**Last Updated:** 2026-11-18  
**Purpose:** Fundamental physics of ultrasonic atomization for C88 design and supplier evaluation

---

## 1. Ultrasonic Atomization Mechanism

### 1.1 Overview

Ultrasonic atomization converts liquid into fine droplets using high-frequency vibration (typically 1-3 MHz). When a piezoelectric ceramic disk vibrates at ultrasonic frequencies, it creates standing capillary waves on the liquid surface. As wave amplitude increases, droplets are ejected from wave crests.

**Key insight:** Droplet size is determined primarily by **frequency**, not power. Higher frequency = smaller droplets.

### 1.2 Two Competing Theories

**Theory 1: Capillary Wave Hypothesis (Lang, 1962)**
- Dominant theory, widely accepted
- Droplets form at peaks of capillary waves
- Wave wavelength determines droplet size
- Works well for MHz frequencies

**Theory 2: Cavitation Hypothesis (Sollner)**
- Bubbles form and collapse, creating shockwaves
- Shockwaves break liquid surface into droplets
- More relevant for low-frequency (<100 kHz) applications
- At MHz frequencies, cavitation requires too much power

**Combined Theory (Bograslavski & Eknadiosyants)**
- Both mechanisms work together
- Cavitation shockwaves enhance capillary wave breaking
- Controversial: some researchers argue cavitation is negligible at MHz

**C88 relevance:** At 1.7MHz, capillary wave mechanism dominates. Cavitation is minimal.

---

## 2. Lang's Equation — Frequency-Droplet Size Relationship

### 2.1 The Fundamental Formula

Lang (1962) derived the relationship between ultrasonic frequency and droplet size:

```
D = 0.34 × λ_c
```

Where:
- D = mean droplet diameter (m)
- λ_c = capillary wavelength (m)
- 0.34 = empirical constant (later adjusted by Yasuda)

### 2.2 Capillary Wavelength

The capillary wavelength is related to frequency by:

```
λ_c = (8πσ / ρf²)^(1/3)
```

Where:
- σ = surface tension of liquid (N/m)
- ρ = liquid density (kg/m³)
- f = ultrasonic frequency (Hz)

### 2.3 Combined Formula

Substituting λ_c into Lang's equation:

```
D = 0.34 × (8πσ / ρf²)^(1/3)
```

**Simplified for water (σ = 0.072 N/m, ρ = 1000 kg/m³):**

```
D ≈ 4.3 × f^(-2/3)    [D in µm, f in MHz]
```

### 2.4 Calculated Values for C88

| Frequency | Droplet Size (water) | Application |
|-----------|---------------------|-------------|
| 1.0 MHz | 4.3 µm | Large droplets, humidification |
| **1.7 MHz** | **2.3 µm** | **C88 primary** |
| 2.0 MHz | 1.9 µm | Medical nebulizers |
| **2.4 MHz** | **1.5 µm** | **C88 alternative** |
| 3.0 MHz | 1.2 µm | Sub-micron applications |
| 5.0 MHz | 0.8 µm | Nanoparticle synthesis |
| 8.5 MHz | 0.5 µm | Sub-micron, pharmaceutical |

**C88 calculation:**
- 1.7 MHz → D = 4.3 × (1.7)^(-2/3) = 4.3 × 0.54 = **2.3 µm** ✓
- 2.4 MHz → D = 4.3 × (2.4)^(-2/3) = 4.3 × 0.43 = **1.8 µm** (not 1.5µm as commonly stated)

**Note:** Actual droplet size distribution is broader than single value. Typical range: 0.5-5 µm for 1.7MHz.

---

## 3. Piezoelectric Effect

### 3.1 Basic Principle

Piezoelectric materials generate electric charge when mechanically stressed (direct effect) and deform when electric field is applied (converse effect).

**Converse piezoelectric effect (used in atomizers):**
```
Strain (S) = d₃₃ × Electric Field (E)
```

Where:
- S = mechanical strain (dimensionless)
- d₃₃ = piezoelectric charge coefficient (C/N or m/V)
- E = applied electric field (V/m)

### 3.2 PZT Materials (Lead Zirconate Titanate)

**Most common piezoelectric ceramic for atomizers:**

| PZT Type | d₃₃ (pC/N) | Curie Temp (°C) | Application |
|----------|------------|-----------------|-------------|
| PZT-4 | 285 | 328 | High-power, industrial |
| PZT-5A | 374 | 365 | General purpose |
| PZT-5H | 593 | 193 | High sensitivity |
| PZT-8 | 248 | 300 | High-power, low loss |

**C88 typical:** PZT-4 or PZT-8 (high power, stable at elevated temperature)

### 3.3 Resonance Frequency

A piezoelectric disk has multiple resonance modes:

**Radial mode (thickness expansion):**
```
f_r = v / (2 × d)
```

Where:
- f_r = radial resonance frequency (Hz)
- v = speed of sound in PZT (~4000 m/s)
- d = disk diameter (m)

**Thickness mode:**
```
f_t = v / (2 × t)
```

Where:
- f_t = thickness resonance frequency (Hz)
- t = disk thickness (m)

**Example calculation:**
- 20mm diameter, 1mm thick PZT disk
- Radial mode: f_r = 4000 / (2 × 0.02) = 100 kHz
- Thickness mode: f_t = 4000 / (2 × 0.001) = 2 MHz

**C88 operation:** 1.7 MHz is between radial and thickness modes. This is the **flexural mode** (bending vibration), which is most efficient for atomization.

---

## 4. Atomization Performance

### 4.1 Mist Production Rate

Mist output scales with:
- Disk diameter (larger = more output)
- Drive power (higher = more output)
- Liquid properties (lower viscosity = easier atomization)

**Empirical formula:**
```
Q = k × P × D²
```

Where:
- Q = mist production rate (g/h)
- P = drive power (W)
- D = disk diameter (mm)
- k = empirical constant (~0.5-1.0 for water)

**Typical performance:**

| Disk Diameter | Drive Power | Mist Rate | Application |
|---------------|-------------|-----------|-------------|
| 16 mm | 10-20 W | 25-35 g/h | Consumer diffuser |
| 20 mm | 20-40 W | 50-100 g/h | Mid-size humidifier |
| 25 mm | 40-100 W | 100-400 g/h | Industrial humidifier |
| 50 mm | 100+ W | 0.5-5 kg/h | Process atomization |

**C88 calculation:**
- 12 piezos × 20mm diameter
- 30W per piezo (typical)
- Total power: 360W
- Expected output: 12 × 75 g/h = **900 g/h** (0.9 L/h)

### 4.2 Efficiency

Atomization efficiency (η) = (energy in mist) / (electrical input)

**Typical values:**
- Consumer diffusers: 30-50%
- Industrial humidifiers: 50-70%
- Medical nebulizers: 70-85%
- **C88 target: >80%**

**Losses:**
- Heat generation in PZT (10-20%)
- Transformer losses (5-10%)
- MOSFET switching losses (5-10%)
- Acoustic radiation (5-10%)

---

## 5. Impedance Characteristics

### 5.1 Equivalent Circuit

Piezoelectric disk can be modeled as:

```
         ┌──[L_m]──[C_m]──[R_m]──┐
         │                        │
    ─────┤                        ├─────
         │                        │
         └────────[C_0]───────────┘
```

Where:
- L_m = motional inductance (mechanical mass)
- C_m = motional capacitance (mechanical compliance)
- R_m = motional resistance (mechanical losses)
- C_0 = static capacitance (electrode capacitance)

### 5.2 Resonance Frequencies

**Series resonance (f_s):**
```
f_s = 1 / (2π √(L_m × C_m))
```
- Minimum impedance
- Maximum current
- Used for atomization

**Parallel resonance (f_p):**
```
f_p = f_s × √(1 + C_m/C_0)
```
- Maximum impedance
- Used for sensing

**Typical values for 20mm, 1.7MHz PZT:**
- f_s = 1.70 MHz
- f_p = 1.75 MHz
- C_0 = 2-4 nF
- R_m = 5-20 Ω

### 5.3 Impedance Matching

Driver must match piezo impedance for maximum power transfer:

```
Z_driver = Z_piezo*
```

Where Z_piezo* is complex conjugate of piezo impedance.

**At resonance:**
- Piezo impedance is mostly resistive (R_m)
- Matching is simple: transformer turns ratio

**Off resonance:**
- Piezo impedance has reactive component
- Requires L-C matching network
- Efficiency drops dramatically

---

## 6. Temperature Effects

### 6.1 Frequency Drift

PZT resonance frequency changes with temperature:

```
Δf/f = k_T × ΔT
```

Where:
- k_T = temperature coefficient (-50 to +50 ppm/°C)
- ΔT = temperature change from 25°C

**Example:**
- k_T = -30 ppm/°C
- ΔT = +25°C (operating at 50°C)
- Δf = 1.7MHz × (-30×10⁻⁶) × 25 = **-1.275 kHz**

**C88 impact:** Frequency drift of ±1kHz is acceptable (1.7MHz ±0.1% spec).

### 6.2 Depolarization

PZT loses piezoelectric properties above Curie temperature:

| PZT Type | Curie Temp (°C) | Max Operating Temp (°C) |
|----------|-----------------|-------------------------|
| PZT-4 | 328 | 200 |
| PZT-5A | 365 | 220 |
| PZT-8 | 300 | 180 |

**C88 operating temp:** 10-50°C (well below Curie temp, no depolarization risk)

### 6.3 Thermal Runaway

At high drive power, PZT self-heating can cause:
1. Temperature rises
2. Mechanical losses increase
3. More heating
4. Thermal runaway → failure

**Prevention:**
- Adequate heat sinking
- Temperature monitoring (NTC thermistor)
- Power derating at high ambient temperature

---

## 7. Power and Drive Requirements

### 7.1 Why High Voltage?

To produce sufficient displacement amplitude (>1µm), the electric field must be high:

```
E = V / t

Where:
- E = electric field (V/m)
- V = applied voltage (V)
- t = piezo thickness (m)
```

**Example:**
- 1mm thick PZT
- Required field: 50,000 V/m
- Required voltage: 50V

**Typical drive voltage:** 50-120V peak-to-peak at piezo element

### 7.2 Why High Frequency?

Higher frequency produces:
- Smaller droplets (Lang equation)
- Faster response time
- Less acoustic noise (above human hearing)

**Trade-offs:**
- Higher frequency = smaller droplets but lower output
- 1.7MHz: good balance for C88 (2.3µm droplets, high output)
- 2.4MHz: smaller droplets (1.8µm) but 30% lower output

### 7.3 Drive Waveform

**Optimal:** Sinusoidal at resonance frequency

**Why not square wave?**
- Square wave contains harmonics (3f, 5f, 7f...)
- Harmonics cause unwanted vibration modes
- Reduced efficiency, more heat

**Practical:** Quasi-sinusoidal (some distortion acceptable)

---

## 8. Key Parameters Summary

### 8.1 For C88 Design

| Parameter | Value | Notes |
|-----------|-------|-------|
| Frequency | 1.7 MHz (primary) | 2.4 MHz alternative |
| Droplet size | 2.3 µm (1.7MHz) | Lang equation |
| Piezo diameter | 20 mm | Standard size |
| Piezo thickness | 1 mm | Typical |
| Drive voltage | 50-120 Vpp | At piezo element |
| Input voltage | 36-48V DC | For driver board |
| Power per piezo | 20-40 W | For 50-100 g/h |
| Total power (12 pcs) | 240-480 W | For 600-1200 g/h |
| Efficiency target | >80% | Including driver losses |

### 8.2 For Supplier Evaluation

**Questions to ask:**
1. What is the resonance frequency tolerance? (±0.05MHz typical)
2. What is the static capacitance C_0? (2-4 nF for 20mm)
3. What is the motional resistance R_m? (5-20 Ω)
4. What is the quality factor Q? (>500 for good piezo)
5. What is the Curie temperature? (>300°C for PZT-4/8)
6. What is the temperature coefficient k_T? (±50 ppm/°C max)

---

## 9. References

### 9.1 Foundational Papers

1. Lang, R.J. (1962). "Ultrasonic atomization of liquids". *Journal of the Acoustical Society of America*, 34(1), 6-8.  
   **Key contribution:** Derived droplet size-frequency relationship

2. Wood, R.W. & Loomis, A.L. (1927). "The physical and biological effects of high-frequency sound-waves of great intensity". *Philosophical Magazine*, 4(19), 417-436.  
   **Key contribution:** First observation of ultrasonic atomization

3. Sollner, K. (1945). "The production of aerosols by the condensation of supersaturated vapours". *Nature*, 155, 41-42.  
   **Key contribution:** Cavitation hypothesis

### 9.2 Technical References

4. "Ultrasonic atomization" — Wikipedia  
   URL: https://en.wikipedia.org/wiki/Ultrasonic_atomization  
   **Key content:** Comprehensive overview, Lang equation derivation

5. "Size distributions of droplets produced by ultrasonic nebulizers" — PMC  
   URL: https://pmc.ncbi.nlm.nih.gov/articles/PMC6468117  
   **Key content:** Experimental validation of Lang equation

6. "Understanding the Mechanism of Piezoelectric Atomizers" — Piezo Direct  
   URL: https://piezodirect.com/understanding-the-mechanism-of-piezoelectric-atomizers/  
   **Key content:** Practical explanation for engineers

### 9.3 Application Notes

7. "Troubleshooting Ultrasonic Atomizer MOSFET Driver Failures" — Industrial Monitor Direct  
   URL: https://industrialmonitordirect.com/blogs/knowledgebase/troubleshooting-ultrasonic-piezo-atomizer-driver-circuit-faults  
   **Key content:** Driver circuit design, failure modes

8. PZT Electronic (Dongguan) — Product specifications  
   URL: https://piezoelements.com  
   **Key content:** Real product specs, driver boards

---

## 10. C88 Application Notes

### 10.1 Why 1.7MHz?

**Droplet size:**
- 1.7MHz → 2.3µm (good for air suspension)
- 2.4MHz → 1.8µm (smaller but lower output)

**Output rate:**
- 1.7MHz → higher output (30% more than 2.4MHz)
- 2.4MHz → lower output but finer droplets

**Supplier availability:**
- 1.7MHz: widely available, proven technology
- 2.4MHz: less common, higher cost

**Recommendation:** Start with 1.7MHz for prototype, evaluate 2.4MHz if droplet size is issue.

### 10.2 Why 12 Piezos?

**Required output:**
- C88 target: 600-1200 g/h (0.6-1.2 L/h)
- Single 20mm piezo: 50-100 g/h
- Required: 6-12 piezos

**Redundancy:**
- 12 piezos allows 2-3 failures before system shutdown
- 6 piezos per module (2 modules) for independent control

**Uniformity:**
- 6 piezos in hexagonal pattern for even distribution
- Avoids hot spots and dry zones

### 10.3 Questions for Saturday SWT Visit

**Technical questions:**
1. What PZT material do you use? (PZT-4, PZT-5A, PZT-8?)
2. What is the d₃₃ coefficient? (should be >300 pC/N)
3. What is the quality factor Q? (should be >500)
4. What is the temperature coefficient? (should be <±50 ppm/°C)
5. How do you ensure frequency consistency across production?

**Performance questions:**
1. What is the atomization efficiency? (target >80%)
2. What is the mist output at 30W drive? (expect 50-100 g/h)
3. What is the frequency tolerance? (±0.05MHz typical)
4. How does output vary with temperature? (10-50°C range)

**Reliability questions:**
1. What is the accelerated aging test protocol?
2. How do you predict 10,000 hour lifetime?
3. What are the failure modes you've observed?
4. What is the warranty policy?

---

**End of Chapter 01 — Piezoelectric Atomization Physics**

**Next:** Chapter 02 — Anti-Corrosion Coating Technologies



<!-- ===== Chapter 02: Anti-Corrosion Coating Technologies / 防腐涂层技术 ===== -->
# Anti-Corrosion Coating Technologies for Piezoelectric Atomizers

**C88 Knowledge Base — Chapter 02**  
**Last Updated:** 2026-11-18  
**Purpose:** Compare coating technologies for C88 piezo atomizers in corrosive disinfection environments

---

## 1. C88 Corrosion Environment

### 1.1 Working Liquids

| Chemical | Concentration | pH | Temperature | Oxidation Potential |
|----------|---------------|-----|-------------|---------------------|
| HOCl (hypochlorous acid) | 650-2,000 ppm | 5.0-6.5 | 10-50°C | Strong (Cl⁻ + OCl⁻) |
| H₂O₂ (hydrogen peroxide) | 7.5% | 2.0-4.5 | 10-50°C | Very strong (•OH radicals) |
| PAA (peracetic acid) | 10,000 ppm (1%) | 1.5-3.0 | 10-50°C | Extreme (CH₃CO₃H) |

### 1.2 Corrosion Mechanisms

1. **Chloride attack** — Cl⁻ ions penetrate coatings, attack metal electrodes
2. **Oxidation** — Strong oxidizers break down organic coatings
3. **Acid hydrolysis** — Low pH degrades glass/ceramic coatings
4. **Cavitation erosion** — Micro-bubble collapse mechanically damages surface
5. **Galvanic corrosion** — Dissimilar metals (Ag electrode + PZT) create electrochemical cells

### 1.3 C88 Severity Rating

**Overall: SEVERE**
- Combined chemical + mechanical attack
- Continuous operation (10,000 hours target)
- No flushing between cycles (concentrated exposure)
- Cavitation accelerates coating breakdown

---

## 2. Glass Glaze Coating (玻璃釉)

### 2.1 Overview

Glass glaze is a vitreous (glass-like) ceramic coating applied to piezoelectric atomizer surfaces. Used by SWT, J&H (吉胜辉), and other Chinese manufacturers.

### 2.2 Composition

**Typical glass glaze formulation:**
- SiO₂ (silicon dioxide): 50-70% — glass former
- B₂O₃ (boron oxide): 10-20% — flux, lowers melting point
- Al₂O₃ (aluminum oxide): 5-15% — stabilizer, chemical resistance
- Na₂O/K₂O (alkali): 5-10% — flux
- ZnO, CaO, BaO: 0-10% — modifiers

**Key properties:**
- Melting point: 500-700°C (during application)
- Coating thickness: 20-50 µm
- Hardness: 5-6 Mohs (scratch resistant)
- Thermal expansion: matched to PZT substrate

### 2.3 Corrosion Resistance Mechanism

**How it works:**
1. **Physical barrier** — Dense, non-porous glass layer blocks liquid penetration
2. **Chemical inertness** — SiO₂/B₂O₃ network resistant to most chemicals
3. **Smooth surface** — Reduces cavitation erosion (fewer nucleation sites)

**Resistance to C88 chemicals:**

| Chemical | Resistance | Mechanism |
|----------|------------|-----------|
| HOCl (pH 5-6.5) | **Good** | Glass stable in mild acid |
| H₂O₂ (pH 2-4.5) | **Moderate** | Acid attacks glass network slowly |
| PAA (pH 1.5-3.0) | **Poor-Moderate** | Strong acid + oxidizer degrades glass |

**Weaknesses:**
- Alkali attack: NaOH/KOH dissolves glass (not relevant for C88)
- HF acid: dissolves glass rapidly (not present in C88)
- Thermal shock: rapid temperature changes can crack glaze

### 2.4 Performance Data

**From SWT (思威特) — confirmed 2026-11-17:**
- Glass glaze atomizer resistant to HOCl, H₂O₂, PAA
- Rated lifetime: 8,000+ hours in disinfection applications
- Frequency: 1.7 MHz, 2.4 MHz available
- No published test data (verbal confirmation only)

**From J&H (吉胜辉) — claimed 2026-11-18:**
- Glass glaze is anti-corrosive
- Technical specifications pending

**Industry benchmarks:**
- Consumer humidifiers: 3,000-5,000 hours (tap water, no corrosion)
- Industrial atomizers: 5,000-8,000 hours (mild chemicals)
- **C88 requirement: 10,000 hours (HOCl/H₂O₂/PAA)**

**Gap analysis:** Glass glaze rated for 8,000h, C88 needs 10,000h. **20% shortfall.**

### 2.5 Advantages and Disadvantages

**Advantages:**
- ✅ Low cost (simple application process)
- ✅ Good chemical resistance (HOCl, mild acids)
- ✅ Smooth surface (reduces cavitation)
- ✅ Proven technology (decades of use)
- ✅ Can be applied to complex geometries

**Disadvantages:**
- ❌ Brittle (cracks under thermal/mechanical shock)
- ❌ Poor resistance to strong acids (pH <2)
- ❌ No published test data for C88 chemicals
- ❌ Lifetime <10,000h in corrosive environments
- ❌ Difficult to inspect for micro-cracks

---

## 3. Electrode Protective Film (电极保护膜) — Audiowell Technology

### 3.1 Overview

Audiowell's proprietary "electrode protective film" technology protects the silver electrodes from corrosion. Specific composition is trade secret, but likely a thin-film ceramic or polymer coating.

### 3.2 Likely Composition

**Based on industry knowledge:**
- **Option A: Parylene (poly-para-xylylene)**
  - Thickness: 1-10 µm
  - Excellent chemical resistance
  - Pinhole-free deposition
  - Used in medical implants

- **Option B: Thin-film ceramic (Al₂O₃, TiO₂)**
  - Thickness: 0.5-2 µm
  - Applied by sputtering or CVD
  - Very hard, scratch resistant
  - Excellent barrier properties

- **Option C: Hybrid organic-inorganic**
  - Sol-gel derived coating
  - Combines flexibility of polymer + hardness of ceramic
  - Thickness: 2-5 µm

### 3.3 Parylene Coating Properties (Best Available Data)

**Parylene C (most common type):**

| Property | Value | Significance |
|----------|-------|--------------|
| Thickness | 1-25 µm | Tunable barrier |
| Dielectric strength | 5,600 V/mil | Excellent insulator |
| Water absorption | <0.1% | Minimal swelling |
| Chemical resistance | Excellent | Resists acids, bases, solvents |
| Pinhole-free | Yes | Complete coverage |
| Biocompatible | Yes | Medical grade |
| Operating temp | -200°C to +80°C | Wide range |

**Chemical resistance (Parylene C):**

| Chemical | Resistance | Notes |
|----------|------------|-------|
| HOCl (pH 5-6.5) | **Excellent** | No degradation reported |
| H₂O₂ (7.5%) | **Excellent** | Resists oxidation |
| PAA (1%) | **Good-Excellent** | Mild resistance to organic acids |
| Saline (0.9% NaCl) | **Excellent** | >30 days without breakdown |

**Comparison to other coatings:**

| Coating | Thickness | Time to breakdown (saline) |
|---------|-----------|---------------------------|
| **Parylene C** | 25 µm | **>30 days** |
| Epoxy | 100 µm | 6 hours |
| Polyurethane | 100 µm | 6 hours |
| Silicone | 75 µm | 58 hours |
| Teflon | 75 µm | 6 hours |

**Key insight:** Parylene is 100x more effective than epoxy/polyurethane at 1/4 the thickness.

### 3.4 Audiowell Claims

**From product documentation:**
- "Electrode protective film technology"
- Acid resistance, alkali resistance, cavitation corrosion resistance
- Only manufacturer that publicly markets rated protective layer
- No published test data (proprietary)

**Assessment:**
- Likely Parylene or similar high-performance coating
- Excellent chemical resistance expected
- No independent verification available

### 3.5 Advantages and Disadvantages

**Advantages:**
- ✅ Superior chemical resistance (HOCl, H₂O₂, PAA)
- ✅ Pinhole-free, complete coverage
- ✅ Thin (1-10 µm) — minimal impact on atomization
- ✅ Flexible — resists cracking under thermal cycling
- ✅ Medical grade (biocompatible)
- ✅ Proven in harsh environments

**Disadvantages:**
- ❌ Expensive (Parylene deposition requires vacuum CVD)
- ❌ Proprietary (Audiowell only, not available to others)
- ❌ No published test data for C88 chemicals
- ❌ Difficult to inspect (transparent coating)
- ❌ Can be damaged by abrasion (soft polymer)

---

## 4. Titanium Film Coating (钛膜) — Xizhe Technology

### 4.1 Overview

Xizhe (西喆) offers "titanium film" atomizers. This likely refers to a titanium dioxide (TiO₂) coating applied by sputtering or CVD.

### 4.2 TiO₂ Coating Properties

**Material properties:**

| Property | Value | Significance |
|----------|-------|--------------|
| Hardness | 6-7 Mohs | Very hard, scratch resistant |
| Density | 4.23 g/cm³ | Dense barrier |
| Melting point | 1,843°C | Thermally stable |
| Thermal expansion | 7-9 ×10⁻⁶ /°C | Matched to PZT |
| Refractive index | 2.4-2.9 | Optical properties |

**Chemical resistance:**

| Chemical | Resistance | Mechanism |
|----------|------------|-----------|
| HOCl | **Excellent** | TiO₂ inert to chlorine |
| H₂O₂ | **Excellent** | TiO₂ stable in peroxide |
| PAA | **Excellent** | TiO₂ resistant to organic acids |
| HF acid | **Poor** | HF dissolves TiO₂ |
| NaOH (strong) | **Moderate** | Alkali attacks TiO₂ slowly |

**Key advantage:** TiO₂ is photocatalytic — under UV light, it can self-clean organic contaminants. Not relevant for C88 (no UV), but indicates chemical stability.

### 4.3 Application Methods

**Sputtering (magnetron sputtering):**
- Thickness: 0.1-2 µm
- Uniform coverage
- Excellent adhesion
- Expensive (vacuum process)

**CVD (chemical vapor deposition):**
- Thickness: 0.5-5 µm
- Conformal coating (covers complex shapes)
- Good adhesion
- Moderate cost

**Sol-gel:**
- Thickness: 1-10 µm
- Low cost (dip coating)
- Lower density than sputtered
- Porous (needs multiple layers)

### 4.4 Performance Data

**From Xizhe (西喆):**
- Titanium film atomizer available
- "More corrosion-resistant than enamel" (marketing claim)
- No published test data
- No lifetime data for C88 chemicals

**Industry benchmarks:**
- TiO₂ coatings used in:
  - Self-cleaning glass (architectural)
  - Dental implants (biocompatible)
  - Solar cells (UV stable)
  - **Not commonly used for ultrasonic atomizers**

**Assessment:** TiO₂ has excellent chemical resistance in theory, but limited track record in ultrasonic atomization.

### 4.5 Advantages and Disadvantages

**Advantages:**
- ✅ Excellent chemical resistance (all C88 chemicals)
- ✅ Very hard (scratch/cavitation resistant)
- ✅ Thermally stable (high melting point)
- ✅ Thin (0.5-2 µm) — minimal impact on atomization
- ✅ Proven in other harsh environments

**Disadvantages:**
- ❌ Expensive (vacuum deposition required)
- ❌ Brittle (can crack under thermal shock)
- ❌ Limited track record in atomization
- ❌ No published C88-specific test data
- ❌ Difficult to apply to complex geometries

---

## 5. Technology Comparison

### 5.1 Summary Table

| Property | Glass Glaze | Electrode Film (Audiowell) | Titanium Film (Xizhe) |
|----------|-------------|----------------------------|----------------------|
| **Thickness** | 20-50 µm | 1-10 µm | 0.5-2 µm |
| **HOCl resistance** | Good | Excellent | Excellent |
| **H₂O₂ resistance** | Moderate | Excellent | Excellent |
| **PAA resistance** | Poor-Moderate | Good-Excellent | Excellent |
| **Hardness** | 5-6 Mohs | Soft (polymer) | 6-7 Mohs |
| **Flexibility** | Brittle | Flexible | Brittle |
| **Cost** | Low | High | Very high |
| **Track record** | Proven (consumer) | Proven (industrial) | Limited |
| **C88 lifetime** | 8,000h (est.) | >10,000h (est.) | Unknown |
| **Availability** | SWT, J&H | Audiowell only | Xizhe only |

### 5.2 C88 Suitability Ranking

**1st: Electrode Protective Film (Audiowell)**
- Best chemical resistance
- Proven in industrial applications
- Likely Parylene or equivalent
- **Risk:** Proprietary, expensive, no independent verification

**2nd: Titanium Film (Xizhe)**
- Excellent chemical resistance (theoretical)
- Very hard, cavitation resistant
- **Risk:** Limited track record, expensive, no C88 data

**3rd: Glass Glaze (SWT, J&H)**
- Good HOCl resistance
- Proven in consumer applications
- Low cost
- **Risk:** Poor PAA resistance, brittle, <10,000h lifetime

### 5.3 Recommendation for C88

**Short-term (prototype):**
- Use **glass glaze** (SWT) — low cost, readily available
- Accept 8,000h lifetime (replace after 1 year)
- Monitor for coating degradation

**Long-term (production):**
- Qualify **Audiowell electrode film** — best performance
- Negotiate technology transfer or licensing
- Or develop in-house Parylene coating capability

**Alternative:**
- Test **titanium film** (Xizhe) — promising but needs validation
- Request 1,000h accelerated aging test with C88 chemicals
- If successful, consider for production

---

## 6. Testing Methodology

### 6.1 Accelerated Aging Test Protocol

**Objective:** Verify coating lifetime in C88 environment

**Test setup:**
- 3 samples per coating type (glass glaze, electrode film, Ti film)
- Immersed in C88 working liquid (HOCl 1000ppm, pH 5.5, 40°C)
- Driven at rated power (30W per piezo)
- Continuous operation

**Measurements (weekly):**
1. **Frequency shift** — target <2% from initial
2. **Current rise** — target <10% from initial
3. **Atomization rate** — target <10% degradation
4. **Visual inspection** — check for cracks, delamination, discoloration
5. **Coating integrity** — dye penetration test for pinholes

**Acceptance criteria:**
- Pass: 1,000h with <2% frequency shift, <10% current rise
- Marginal: 500-1,000h with 2-5% frequency shift
- Fail: <500h or >5% frequency shift

**Acceleration factor:**
- 1,000h test @ 40°C ≈ 5,000h real-world @ 25°C (Arrhenius model)
- To verify 10,000h lifetime: need 2,000h test OR higher temperature (60°C)

### 6.2 Chemical Resistance Test

**Objective:** Verify coating resistance to HOCl, H₂O₂, PAA

**Test setup:**
- 3 samples per coating per chemical
- Immersed for 1,000h at 40°C
- Chemicals: HOCl 2000ppm (pH 5.5), H₂O₂ 7.5% (pH 3.0), PAA 1% (pH 2.5)

**Measurements:**
1. **Weight change** — target <1% (indicates dissolution)
2. **Thickness change** — target <5% (indicates erosion)
3. **Adhesion test** — tape pull-off (ASTM D3359)
4. **SEM inspection** — check for micro-cracks, delamination

**Acceptance criteria:**
- Pass: <1% weight change, <5% thickness change, no delamination
- Marginal: 1-3% weight change, 5-10% thickness change
- Fail: >3% weight change or visible damage

---

## 7. Questions for Saturday SWT Visit

### 7.1 Glass Glaze Specific Questions

1. **Composition:** What is the glass glaze composition? (SiO₂/B₂O₃/Al₂O₃ ratios?)
2. **Thickness:** What is the coating thickness? (20µm? 50µm?)
3. **Application method:** How is it applied? (spray, dip, screen print?)
4. **Firing temperature:** What temperature is it fired at? (500°C? 700°C?)
5. **Adhesion:** How do you ensure adhesion to PZT? (surface treatment?)

### 7.2 Chemical Resistance Questions

1. **Test data:** Do you have test data for HOCl 2000ppm immersion?
2. **Lifetime:** What is the rated lifetime in HOCl? (8,000h? 10,000h?)
3. **PAA resistance:** How does glass glaze perform in PAA 1% (pH 2.5)?
4. **Failure mode:** How does the coating fail? (cracking? dissolution? delamination?)
5. **Inspection:** How do you inspect for coating damage? (visual? dye test?)

### 7.3 Comparison Questions

1. **vs Audiowell:** How does your glass glaze compare to Audiowell's electrode film?
2. **vs titanium film:** Have you considered titanium film coating? Why not?
3. **Cost:** What is the cost difference between glass glaze and other coatings?
4. **Customization:** Can you apply Parylene coating if we provide it?

### 7.4 Validation Questions

1. **Sample availability:** Can you provide 3 samples for our testing?
2. **Test protocol:** Will you witness our 1,000h aging test?
3. **Warranty:** What is your warranty policy for coating failure?
4. **Quality control:** How do you ensure coating quality in production?

---

## 8. References

### 8.1 Technical References

1. "High Frequency Glass Glaze Atomizing Disc" — B2B Wiki  
   URL: https://b2bwiki.aipage.com/en/encyclopedic/high-frequency-glass-glaze-atomizing-disc  
   **Key content:** Glass glaze composition, properties, applications

2. "Parylene Coating Properties" — HZO Inc.  
   URL: https://hzo.com/coatings/parylene-coating/properties/  
   **Key content:** Parylene electrical, mechanical, chemical properties

3. "A Selective Review of Ceramic, Glass and Glass–Ceramic Protective Coatings" — PMC  
   URL: https://pmc.ncbi.nlm.nih.gov/articles/PMC10253744/  
   **Key content:** Glass/ceramic coatings for protection

4. "Attack on Glass: Corrosion Attack Mechanisms" — Digital Fire  
   URL: https://digitalfire.com/article/attack+on+glass%3A+corrosion+attack+mechanisms  
   **Key content:** Glass corrosion mechanisms, pH effects

### 8.2 Supplier Documents

5. SWT (思威特) — Product catalog and verbal confirmation (2026-11-17)  
   **Key content:** Glass glaze atomizer, anti-corrosion claims

6. Audiowell (奥迪威) — Product documentation  
   URL: https://audiowell.com/transducer.html  
   **Key content:** Electrode protective film technology

7. Xizhe (西喆) — Product specifications  
   URL: http://www.xzpzt.com  
   **Key content:** Titanium film atomizer

### 8.3 Standards

8. ASTM D3359 — Standard Test Methods for Measuring Adhesion by Tape Test  
9. IEC 60068 — Environmental Testing (temperature, humidity, vibration)  
10. ISO 9227 — Corrosion Testing by Salt Spray Methods

---

## 9. C88 Application Notes

### 9.1 Key Findings

1. **Glass glaze is NOT sufficient for 10,000h C88 lifetime**
   - Rated for 8,000h in best case
   - Poor PAA resistance (pH 1.5-3.0)
   - Brittle, cracks under thermal cycling

2. **Audiowell electrode film is likely the best solution**
   - Probably Parylene or equivalent
   - Excellent chemical resistance
   - Proven in industrial applications
   - **Problem:** Proprietary, expensive

3. **Titanium film is promising but unproven**
   - Excellent chemical resistance (theoretical)
   - Very hard, cavitation resistant
   - **Problem:** No atomization track record

### 9.2 Recommended Strategy

**Phase 1 (Prototype — Now):**
- Use SWT glass glaze atomizers (available, low cost)
- Accept 8,000h lifetime (replace after 1 year)
- Run 1,000h aging test to validate

**Phase 2 (Validation — Month 3-6):**
- Qualify Audiowell electrode film modules
- Negotiate pricing for 100+ units
- Run 2,000h accelerated aging test

**Phase 3 (Production — Month 6+):**
- Switch to Audiowell modules (if qualified)
- Or develop in-house Parylene coating
- Target 10,000h+ lifetime

### 9.3 Risk Mitigation

**Risk 1: Glass glaze fails early**
- Mitigation: Keep spare atomizers in stock
- Replacement cost: ~$10-20 per piezo
- Downtime: 30 minutes (swap modules)

**Risk 2: Audiowell won't share technology**
- Mitigation: Develop alternative (Parylene coating service)
- Vendors: HZO, Specialty Coating Systems (SCS)
- Cost: ~$50-100 per piezo (coating service)

**Risk 3: No coating lasts 10,000h**
- Mitigation: Design for easy replacement
- Modular design (6 piezos per module)
- Scheduled maintenance (replace every 8,000h)

---

**End of Chapter 02 — Anti-Corrosion Coating Technologies**

**Next:** Chapter 03 — Driver Circuit Design (completed separately)



<!-- ===== Chapter 03: Driver Circuit Design / 驱动电路设计 ===== -->
# Piezoelectric Atomizer Driver Circuit Design

**C88 Knowledge Base — Chapter 03**  
**Last Updated:** 2026-11-18  
**Purpose:** Technical reference for C88 piezo driver design and supplier evaluation

---

## 1. Why Piezo Needs 36-48V AC Drive Voltage

### 1.1 The Voltage-Amplitude Relationship

Piezoelectric ceramic disks convert electrical energy to mechanical vibration. The displacement amplitude (δ) is proportional to the applied electric field (E):

```
δ = d₃₃ × V × (t/t) = d₃₃ × V
```

Where:
- d₃₃ = piezoelectric coefficient (typically 200-650 pC/N for PZT)
- V = applied voltage
- t = piezo thickness

For a 20mm diameter, 1mm thick PZT disk to produce sufficient capillary wave amplitude (>1µm) for atomization, the required drive voltage is typically **50-120V peak-to-peak** at the piezo element.

### 1.2 Step-Up Transformer Architecture

Most commercial drivers use a **step-up transformer** to convert low-voltage DC (12-48V) to high-voltage AC:

```
DC Input (12-48V) → MOSFET Switch → Transformer Primary → Step-Up → Piezo (50-120Vpp)
```

**Why 35-48V input?**
- Lower voltage (12V) requires higher current for same power (P = V×I)
- Higher voltage reduces I²R losses in MOSFET and wiring
- 35V is common for 25mm atomizers; 48V for industrial units
- C88 spec: 36-48V DC allows flexibility in power supply selection

### 1.3 Real Product Example

**PZT Electronic PZT-25-1.7BU Board:**
- Input: 35V DC
- Frequency: 1.7 MHz
- Power: 35W
- Atomizer diameter: 25mm
- Water level: 25-30mm
- Includes: atomization board + power board + reed switch + indicator

---

## 2. Driver Circuit Topologies

### 2.1 Self-Oscillating Circuit (Colpitts/Pierce)

**Simplest approach:** Piezo element itself forms part of the oscillator feedback network.

```
         +Vcc
          |
         [L]  (choke)
          |
    +-----+-----+
    |           |
   [C1]       [C2]
    |           |
    +-----+-----+
          |
         [Piezo]
          |
         GND
```

**Advantages:**
- No external oscillator needed
- Automatically tracks piezo resonance
- Low component count
- Low cost (<$5 BOM)

**Disadvantages:**
- Frequency stability ±0.5-1%
- Hard to control power output
- Not suitable for multi-piezo arrays

**Use case:** Consumer humidifiers, aroma diffusers

### 2.2 Separate Oscillator + MOSFET Driver

**Modern approach:** Dedicated oscillator IC drives MOSFET gate.

```
Crystal Oscillator (1.7MHz) → Gate Driver IC → MOSFET → Transformer → Piezo
```

**Advantages:**
- Precise frequency control (±10 ppm with TCXO)
- PWM power control possible
- Can synchronize multiple drivers
- Better EMI performance

**Disadvantages:**
- Higher BOM cost ($10-20)
- More complex PCB layout
- Requires impedance matching network

**Use case:** Industrial atomizers, medical nebulizers, C88

### 2.3 Half-Bridge / Push-Pull Topology

**High-power approach:** Two MOSFETs alternate to drive transformer primary.

```
         +Vcc
          |
    +-----+-----+
    |           |
  [Q1]       [Q2]
    |           |
    +-----+-----+
          |
     [Transformer Primary]
          |
         GND
```

**Advantages:**
- Doubles power output vs single-ended
- Lower MOSFET stress (each sees Vcc, not 2×Vcc)
- Better efficiency (70-90%)
- Suitable for multi-piezo arrays

**Disadvantages:**
- Requires complementary gate drive (high-side + low-side)
- Risk of shoot-through if timing is wrong
- More complex control logic

**Use case:** C88 (12 piezos), industrial humidifiers

---

## 3. Frequency Control Methods

### 3.1 Crystal Oscillator (XO)

**Basic:** AT-cut quartz crystal, fixed frequency.

- Accuracy: ±20-50 ppm
- Stability: ±0.01%/°C
- Cost: $0.50-2.00
- **C88 suitability:** Good for fixed 1.7MHz design

### 3.2 Temperature-Compensated Crystal (TCXO)

**Enhanced:** Internal compensation for temperature drift.

- Accuracy: ±0.5-2.5 ppm
- Stability: ±0.001%/°C
- Cost: $2-10
- **C88 suitability:** Best for precision atomization

### 3.3 Phase-Locked Loop (PLL)

**Flexible:** Locks to reference frequency, can be tuned.

- Accuracy: ±0.1 ppm (locked)
- Tuning range: ±0.1-1%
- Cost: $3-8
- **C88 suitability:** Good for frequency sweep testing

### 3.4 Direct Digital Synthesis (DDS)

**Most flexible:** Digital generation, programmable frequency.

- Accuracy: <0.1 ppm
- Tuning range: Wide (DC to MHz)
- Cost: $5-15
- **C88 suitability:** Best for R&D, multi-frequency testing

**Recommendation for C88:** TCXO for production (cost-effective, stable), DDS for prototype (flexible).

---

## 4. Impedance Matching Network

### 4.1 Why Matching Matters

Piezo element impedance at resonance:
- Capacitive reactance: Xc = 1/(2πfC) ≈ 50-200Ω
- Resistive component: R ≈ 5-20Ω (motional resistance)

Without matching, most power reflects back to driver, causing:
- Low atomization efficiency
- MOSFET overheating
- Frequency instability

### 4.2 L-Network Matching

**Simplest:** Series inductor cancels capacitive reactance.

```
Driver → [L] → [Piezo] → GND
```

Where L is chosen so: 2πfL = 1/(2πfC)

**Example calculation:**
- Piezo capacitance: C = 2nF
- Frequency: f = 1.7MHz
- Required inductance: L = 1/((2π×1.7MHz)²×2nF) = 4.4µH

### 4.3 Transformer Matching

**Most common:** Step-up transformer provides both voltage boost and impedance transformation.

```
Driver (low-Z) → [Transformer] → Piezo (high-Z)
```

Turns ratio: n = √(Z_piezo / Z_driver)

**Example:**
- Driver output impedance: 5Ω
- Piezo impedance at resonance: 100Ω
- Turns ratio: n = √(100/5) = 4.5:1

### 4.4 Pi-Network (Advanced)

**Best performance:** L-C-L network for broadband matching.

```
Driver → [L1] → [C] → [L2] → Piezo
```

**Advantages:**
- Wider bandwidth
- Better harmonic suppression
- Can provide DC isolation

**Use case:** High-power industrial atomizers

---

## 5. Power Amplification Stages

### 5.1 MOSFET Selection Criteria

For 1.7MHz, 35W atomizer driver:

| Parameter | Minimum Spec | Recommended |
|-----------|--------------|-------------|
| V_DS (drain-source) | ≥60V | 100V |
| I_D (continuous drain) | ≥3A | 5A |
| R_DS(on) | <100mΩ | <50mΩ |
| Q_g (gate charge) | <20nC | <10nC |
| f_T (transition freq) | ≥50MHz | ≥100MHz |

**Recommended parts:**
- IRF510 (100V, 5A, R_DS=0.54Ω) — budget choice
- IRF740 (400V, 10A, R_DS=0.55Ω) — overkill but robust
- CSD17571Q5A (30V, 10A, R_DS=3.5mΩ) — low-voltage high-efficiency

### 5.2 Gate Drive Requirements

At 1.7MHz, gate drive is critical:

- Gate voltage: 10-15V (logic-level MOSFETs: 4.5V)
- Gate current: I_g = Q_g × f = 10nC × 1.7MHz = 17mA
- Rise time: <50ns (to minimize switching losses)
- Drive IC: TC4420 (high-speed, 6A peak)

### 5.3 Thermal Management

MOSFET power dissipation:

```
P_loss = I²_RMS × R_DS(on) + switching losses

Example:
- I_RMS = 2A
- R_DS(on) = 50mΩ
- Conduction loss = 2² × 0.05 = 0.2W
- Switching loss @ 1.7MHz ≈ 0.5W
- Total ≈ 0.7W
```

**Heatsink requirement:**
- Thermal resistance: θ_SA < (T_J_max - T_A) / P_loss
- For T_J_max = 150°C, T_A = 40°C, P_loss = 0.7W
- θ_SA < 157°C/W → small clip-on heatsink sufficient

---

## 6. Temperature Compensation

### 6.1 Piezo Frequency Drift

PZT resonance frequency shifts with temperature:

```
Δf/f = k_T × ΔT

Where:
- k_T = temperature coefficient ≈ -50 to +50 ppm/°C
- ΔT = temperature change from 25°C
```

**Example:**
- k_T = -30 ppm/°C
- ΔT = +25°C (operating at 50°C)
- Δf = 1.7MHz × (-30×10⁻⁶) × 25 = -1.275kHz

### 6.2 Compensation Methods

**Method 1: TCXO**
- Use temperature-compensated crystal oscillator
- Drift: ±0.5 ppm/°C
- Cost: $3-8

**Method 2: Frequency Tracking**
- Monitor piezo current phase
- Adjust oscillator to maintain resonance
- Requires PLL or DDS

**Method 3: NTC Thermistor**
- Measure piezo temperature
- Adjust oscillator frequency in software
- Requires MCU with lookup table

**C88 recommendation:** TCXO for simplicity (±0.5kHz drift acceptable for 1.7MHz ±1% spec).

---

## 7. Complete Module Design

### 7.1 Audiowell UM0108-003

**Specifications:**
- Frequency: 1.7 ±0.05 MHz
- Current: 0.65A
- Atomization rate: ≥360 mL/h
- Operating temp: +3°C to +50°C
- Input voltage: 24V DC (internal step-up)
- Control: Digital (UART/I2C)
- Dimensions: 45×45×25mm

**Features:**
- Integrated driver + piezo + housing
- Temperature sensor (NTC)
- Water level detection
- Over-current protection
- LED status indicator

**Price:** ~$25-40/unit (MOQ 100+)

### 7.2 KERI Integrated Module

**Specifications:**
- Frequency: 1.7 MHz or 2.4 MHz
- Power: 20-40W
- Anti-corrosion: HOCl/H₂O₂ compatible
- Lifetime: ≥8,000 hours
- Input voltage: 35V DC
- Control: PWM or analog

**Features:**
- Glass glaze coated piezo
- Stainless steel housing (316L)
- IP68 waterproof
- Dry-run protection

**Price:** ~$30-50/unit (MOQ 50+)

### 7.3 C88 Custom Module Design

**Requirements:**
- 6 piezos per module (2 modules = 12 total)
- 1.7 MHz or 2.4 MHz (TBD)
- HOCl/H₂O₂/PAA compatible
- Lifetime ≥10,000 hours
- 316L housing with Parylene coating

**Proposed architecture:**
```
48V DC Input → Half-Bridge Driver → 6× Transformer Secondaries → 6× Piezos
                ↓
           TCXO (1.7MHz) + MCU (STM32)
                ↓
        Current sensing + temp monitoring
```

**Estimated BOM:**
- Half-bridge driver: $8
- TCXO: $5
- MCU (STM32F103): $3
- 6× transformers: $12
- Current sensors: $6
- PCB + passives: $10
- **Total: ~$44** (vs $25-40 for Audiowell module)

**Trade-off:** Custom module is more expensive but gives full control and anti-corrosion design.

---

## 8. Design Trade-offs

### 8.1 Cost vs Performance

| Approach | Cost | Performance | Flexibility |
|----------|------|-------------|-------------|
| Buy Audiowell module | $25-40 | Good (proven) | Low (fixed) |
| Buy KERI module | $30-50 | Good (anti-corrosion) | Low (fixed) |
| Custom design | $44+ | Best (optimized) | High (customizable) |

### 8.2 Single Driver vs Multi-Channel

**Single driver (all 12 piezos in parallel):**
- Pros: Simple, low cost
- Cons: One failure kills all, hard to balance

**Multi-channel (6 drivers, 2 piezos each):**
- Pros: Redundancy, individual control
- Cons: Higher cost, complex PCB

**C88 recommendation:** 2 channels (6 piezos each), one per module.

### 8.3 Fixed Frequency vs Tunable

**Fixed (TCXO):**
- Pros: Simple, stable, low cost
- Cons: Can't optimize for piezo variation

**Tunable (DDS/PLL):**
- Pros: Can track piezo resonance, multi-frequency
- Cons: Complex, expensive, EMI issues

**C88 recommendation:** Fixed TCXO for production, DDS for R&D prototype.

---

## 9. Cost Considerations

### 9.1 Driver Board BOM Breakdown

**Low-cost design (single channel, 6 piezos):**

| Component | Qty | Unit Cost | Total |
|-----------|-----|-----------|-------|
| Half-bridge MOSFETs | 2 | $1.50 | $3.00 |
| Gate driver IC | 1 | $2.00 | $2.00 |
| TCXO 1.7MHz | 1 | $4.00 | $4.00 |
| Step-up transformer | 6 | $2.00 | $12.00 |
| Current sense resistors | 6 | $0.20 | $1.20 |
| MCU (STM32F103) | 1 | $3.00 | $3.00 |
| PCB (2-layer, 100×80mm) | 1 | $5.00 | $5.00 |
| Passives (caps, resistors) | 1 lot | $3.00 | $3.00 |
| Connectors | 1 lot | $2.00 | $2.00 |
| **Total** | | | **$35.20** |

### 9.2 Volume Pricing

At 100-unit volume:
- PCB: $3.50 (vs $5.00 prototype)
- Transformers: $1.50 ea (vs $2.00)
- **Total: ~$30/unit**

At 1000-unit volume:
- PCB: $2.00
- Transformers: $1.00 ea
- **Total: ~$25/unit**

### 9.3 Buy vs Build Analysis

| Option | Unit Cost | NRE | Lead Time | Risk |
|--------|-----------|-----|-----------|------|
| Audiowell module | $35 | $0 | 4 weeks | Low |
| KERI module | $40 | $0 | 6 weeks | Low |
| Custom design | $25 | $5,000 | 12 weeks | Medium |

**Break-even:** Custom design pays off at ~200 units (saves $10-15/unit vs modules).

**C88 recommendation:** Start with Audiowell/KERI modules for prototype (10-50 units), switch to custom design for production (100+ units).

---

## 10. References

### 10.1 Technical Papers

1. "Switching mode power convertor with phase locked oscillator as a driver for an ultrasonic atomizer" — IEEE, 2014  
   URL: https://ieeexplore.ieee.org/document/7070972

2. "Ultrasonic nebulizer for producing high-volume sub-micron droplets" — US Patent 7,129,619  
   URL: https://patents.google.com/patent/US7129619B2/en

3. "Ultrasonic Wave Nebulizer Driving Circuit" — US Patent 4,359,697  
   URL: https://patents.searchlight.law/doc/US4359697

### 10.2 Application Notes

4. "Troubleshooting Ultrasonic Atomizer MOSFET Driver Failures" — Industrial Monitor Direct  
   URL: https://industrialmonitordirect.com/blogs/knowledgebase/troubleshooting-ultrasonic-piezo-atomizer-driver-circuit-faults

5. "35V Atomizer Power Mainboard" — PZT Electronic (Dongguan)  
   URL: https://piezoelements.com/pcb-driver-board/35v-atomizer-power-mainboard.html

### 10.3 Supplier Documents

6. Audiowell UM0108-003 Smart Atomizer Module Datasheet  
   URL: https://audiowell.com/transducer.html

7. KERI Anti-Corrosion Atomizer Module Specifications  
   URL: https://www.keri.com.cn (requires login)

### 10.4 Design Tools

8. "2.4mhz Piezo atomizer circuit" — Vancouver Hack Space forum discussion  
   URL: https://talk.vanhack.ca/t/2-4mhz-piezo-atomizer-circuit/7420

---

## 11. C88 Application Notes

### 11.1 Key Design Decisions

**Frequency:** 1.7MHz (primary) or 2.4MHz (alternative)
- 1.7MHz: Larger droplets (2.3µm), higher output, proven technology
- 2.4MHz: Smaller droplets (1.5µm), lower output, better penetration

**Voltage:** 36-48V DC
- 36V: Lower stress on components, suitable for prototype
- 48V: Higher power, better for production (12 piezos)

**Module configuration:** 2 modules × 6 piezos
- Independent channels for redundancy
- Each channel: 35W driver, 6 transformers

### 11.2 Questions for Supplier Discussions

**For Audiowell:**
1. Can UM0108 module be customized for HOCl environment?
2. What is the input voltage range (24V or 48V)?
3. Can modules be synchronized (phase-locked)?
4. What is the MTBF (mean time between failures)?

**For KERI:**
1. Glass glaze coating thickness and composition?
2. Accelerated aging test data for HOCl 2000ppm?
3. Can driver board be separated from piezo housing?
4. What is the frequency tolerance (±0.05MHz or tighter)?

**For J&H (吉胜辉):**
1. Glass glaze anti-corrosion mechanism?
2. Test data for H₂O₂ 7.5% and PAA 10,000ppm?
3. Can you provide complete module (piezo + driver)?
4. What is the minimum order quantity for custom design?

### 11.3 Prototype Testing Plan

**Phase 1: Single piezo validation (Week 1-2)**
- Test 1× piezo with commercial driver board
- Measure atomization rate, droplet size, power consumption
- Validate frequency and voltage requirements

**Phase 2: 6-piezo module (Week 3-4)**
- Build custom driver board for 6 piezos
- Test uniformity (all 6 piezos same output?)
- Measure thermal performance (MOSFET temperature)

**Phase 3: Full system (Week 5-6)**
- Integrate 2 modules (12 piezos total)
- Test in 316L tank with HOCl solution
- Measure lifetime (accelerated aging)

---

**End of Chapter 03 — Driver Circuit Design**

**Next:** Chapter 04 — Lifetime Testing Methods (in progress)



<!-- ===== Chapter 04: Lifetime Testing Methods / 寿命测试方法 ===== -->
# Piezoelectric Atomizer Lifetime Testing Methods

**C88 Knowledge Base — Chapter 04**  
**Last Updated:** 2026-11-18  
**Purpose:** Methods to verify piezo atomizer lifetime claims for C88 (target: 10,000 hours)

---

## 1. Lifetime Testing Challenge

### 1.1 The Problem

C88 requires **10,000 hours** (3-5 years at 6-8h/day) lifetime. Testing this directly would take:
- 10,000 hours = **417 days** = **14 months** of continuous operation
- Not practical for product development timeline

### 1.2 The Solution: Accelerated Aging

Use **accelerated stress testing** to simulate years of operation in weeks/months:
- Increase temperature (Arrhenius model)
- Increase chemical concentration
- Increase drive power
- Apply thermal cycling

**Goal:** Compress 10,000h into 500-2,000h test time.

---

## 2. Accelerated Aging Models

### 2.1 Arrhenius Model (Temperature Acceleration)

**Fundamental equation:**
```
AF = exp[(Ea/k) × (1/T_use - 1/T_test)]
```

Where:
- AF = acceleration factor
- Ea = activation energy (eV) — typically 0.3-1.0 eV for piezo degradation
- k = Boltzmann constant (8.617×10⁻⁵ eV/K)
- T_use = use temperature (K) — e.g., 298K (25°C)
- T_test = test temperature (K) — e.g., 333K (60°C)

**Example calculation:**
- Ea = 0.7 eV (typical for PZT degradation)
- T_use = 25°C = 298K
- T_test = 60°C = 333K
- AF = exp[(0.7/8.617×10⁻⁵) × (1/298 - 1/333)]
- AF = exp[8123 × (0.003356 - 0.003003)]
- AF = exp[8123 × 0.000353]
- AF = exp[2.87] = **17.6**

**Interpretation:** Testing at 60°C accelerates aging by 17.6×. So 500h test = 8,800h real-world.

### 2.2 Eyring Model (Combined Stress)

**More comprehensive:**
```
AF = (T_test/T_use)^n × exp[(Ea/k) × (1/T_use - 1/T_test)]
```

Where:
- n = temperature exponent (typically 2-4 for piezo)
- Other factors: humidity, voltage, chemical concentration

**For C88:**
- Temperature: 25°C → 60°C (AF = 17.6)
- HOCl concentration: 650ppm → 2000ppm (AF = 3.1)
- Drive power: 30W → 45W (AF = 1.5)
- **Combined AF = 17.6 × 3.1 × 1.5 = 82**

**Interpretation:** 122h test @ 60°C, 2000ppm HOCl, 45W = 10,000h real-world.

### 2.3 Inverse Power Law (Voltage/Power Acceleration)

**For electrical stress:**
```
AF = (V_test/V_use)^n
```

Where:
- n = voltage exponent (typically 2-4 for piezo)
- V = drive voltage or power

**Example:**
- V_use = 30W, V_test = 45W
- n = 3 (typical)
- AF = (45/30)³ = 1.5³ = **3.4**

---

## 3. Standard Test Protocols

### 3.1 IEC 60068 (Environmental Testing)

**Relevant parts:**
- IEC 60068-2-1: Cold testing
- IEC 60068-2-2: Dry heat testing
- IEC 60068-2-6: Vibration testing
- IEC 60068-2-30: Damp heat cycling

**Not directly applicable:** IEC 60068 doesn't cover ultrasonic atomizers specifically.

### 3.2 ASTM Standards

**Relevant standards:**
- ASTM B117: Salt spray (fog) testing
- ASTM G85: Modified salt spray testing
- ASTM D3359: Adhesion testing (tape test)

**Not directly applicable:** No ASTM standard for piezo atomizer lifetime.

### 3.3 Industry Practice (No Formal Standard)

**Typical protocol for ultrasonic atomizers:**

**Test setup:**
- 5-10 samples per batch
- Immersed in working liquid (HOCl, H₂O₂, or PAA)
- Driven at rated power (continuous or duty cycle)
- Temperature controlled (25°C, 40°C, or 60°C)

**Measurements (weekly or bi-weekly):**
1. Resonance frequency (impedance analyzer)
2. Static capacitance (LCR meter)
3. Drive current (oscilloscope + current probe)
4. Atomization rate (weigh method)
5. Visual inspection (microscope)

**Pass/fail criteria:**
- Frequency shift <2% from initial
- Capacitance change <5% from initial
- Current rise <10% from initial
- Atomization rate degradation <10%
- No visible coating damage, cracks, or delamination

**Test duration:**
- 500h: Quick screen (reject bad designs)
- 1,000h: Standard qualification
- 2,000h: High-reliability qualification
- 5,000h: Extended life verification

---

## 4. C88-Specific Test Protocol

### 4.1 Test Matrix

**Phase 1: Quick Screen (500h)**
- 3 samples per supplier (SWT, Audiowell, J&H)
- HOCl 1000ppm, 40°C, 30W continuous
- Weekly measurements
- **Goal:** Eliminate obviously bad designs

**Phase 2: Standard Qualification (1,000h)**
- 5 samples per supplier (top 2 from Phase 1)
- HOCl 2000ppm, 50°C, 40W continuous
- Bi-weekly measurements
- **Goal:** Verify 10,000h equivalent (AF=10)

**Phase 3: High-Reliability (2,000h)**
- 3 samples from winning supplier
- Combined stress: HOCl 2000ppm + H₂O₂ 7.5% + PAA 1%, 60°C, 45W
- Weekly measurements + SEM inspection at end
- **Goal:** Verify 10,000h equivalent (AF=5)

### 4.2 Test Equipment

**Required:**
1. **Temperature chamber** — capable of 60°C ±1°C, 10L volume
2. **Chemical bath** — glass or PTFE tank, 5L volume
3. **Driver boards** — 12 channels (one per piezo), constant power
4. **Impedance analyzer** — Agilent 4294A or equivalent
5. **LCR meter** — for capacitance measurement
6. **Oscilloscope + current probe** — for drive current
7. **Precision balance** — 0.1mg resolution (for atomization rate)
8. **Microscope** — 50-200× for visual inspection

**Estimated cost:**
- Temperature chamber: ¥15,000-30,000
- Chemical bath: ¥2,000-5,000
- Driver boards: ¥5,000-10,000
- Impedance analyzer: ¥30,000-50,000 (rent or borrow)
- Other instruments: ¥10,000-20,000
- **Total: ¥62,000-115,000** ($8,500-16,000)

### 4.3 Test Procedure

**Step 1: Initial characterization (Day 0)**
- Measure all samples: frequency, capacitance, current, atomization rate
- Photograph each sample (microscope)
- Record initial values in test log

**Step 2: Start aging test**
- Place samples in chemical bath
- Set temperature to target (40°C or 60°C)
- Start driver boards (continuous operation)
- Log temperature, current, voltage (data logger)

**Step 3: Weekly measurements**
- Remove samples from bath (briefly)
- Rinse with DI water
- Measure frequency, capacitance, current, atomization rate
- Visual inspection (microscope)
- Return to bath
- Record all data

**Step 4: Intermediate inspection (500h, 1000h)**
- Full visual inspection (microscope + SEM if available)
- Dye penetration test (check for coating pinholes)
- Adhesion test (tape pull-off, destructive — use spare samples)
- Cross-section analysis (if failure observed)

**Step 5: Final analysis (end of test)**
- All measurements from Step 3
- SEM/EDS analysis of coating (if available)
- Compare to initial values
- Calculate degradation rates
- Determine pass/fail

### 4.4 Data Analysis

**Degradation rate calculation:**
```
Degradation rate = (Final - Initial) / Test duration

Example:
- Initial frequency: 1.700 MHz
- Final frequency (1000h): 1.720 MHz
- Degradation rate: (1.720 - 1.700) / 1000 = 0.020 MHz/1000h = 1.18%/1000h
```

**Lifetime extrapolation:**
```
Time to 2% shift = 2% / (1.18%/1000h) = 1,695h (test)
Equivalent real-world = 1,695h × AF = 1,695 × 10 = 16,950h ✓
```

**Weibull analysis (if multiple samples):**
- Plot failure probability vs time
- Fit Weibull distribution
- Extract characteristic life (η) and shape parameter (β)
- Predict B10 life (10% failure rate)

---

## 5. Supplier Claims Verification

### 5.1 What Suppliers Claim

| Supplier | Claimed Lifetime | Test Conditions | Verification |
|----------|-----------------|-----------------|--------------|
| SWT | 8,000h | "Disinfection application" | Verbal only, no data |
| Audiowell | 10,000h+ | "Industrial environment" | No data published |
| J&H | Unknown | Unknown | Pending specs |
| KERI | 8,000h | "HOCl compatible" | Datasheet available |
| APC (US) | 10,000h | "Standard nebulizer" | Datasheet available |

### 5.2 Red Flags

**Claims to question:**
1. "10,000 hours" with no test data — likely marketing
2. "HOCl compatible" with no concentration/pH specified — vague
3. "Industrial grade" with no standard referenced — meaningless
4. "Long lifetime" with no acceleration factor — unverifiable

**Good signs:**
1. Test report with specific conditions (temp, concentration, power)
2. Acceleration factor calculated (Arrhenius model)
3. Sample size stated (n≥5 for statistical validity)
4. Pass/fail criteria defined (frequency shift, current rise)
5. Third-party lab verification (SGS, CSTC)

### 5.3 Questions for Suppliers

**For SWT (Saturday visit):**
1. What is your test protocol for 8,000h claim?
2. What acceleration factor did you use?
3. How many samples were tested?
4. What were the pass/fail criteria?
5. Can you provide the test report?
6. Will you allow us to witness a 1,000h test?

**For Audiowell:**
1. What is the test protocol for "10,000h+" claim?
2. What chemicals were tested (HOCl, H₂O₂, PAA)?
3. What temperature and concentration?
4. Can you provide the test report?
5. Will you support our qualification testing?

**For KERI:**
1. Your datasheet shows 8,000h — what are the test conditions?
2. Can you provide the full test report?
3. Will you customize for C88 chemicals (HOCl 2000ppm)?
4. What is the acceleration factor?

---

## 6. Acceptance Criteria for C88

### 6.1 Pass/Fail Criteria

**After 1,000h test (HOCl 2000ppm, 50°C, 40W):**

| Parameter | Pass | Marginal | Fail |
|-----------|------|----------|------|
| Frequency shift | <2% | 2-5% | >5% |
| Capacitance change | <5% | 5-10% | >10% |
| Current rise | <10% | 10-20% | >20% |
| Atomization rate | >90% initial | 80-90% | <80% |
| Visual inspection | No damage | Minor discoloration | Cracks, delamination |

### 6.2 Lifetime Prediction

**If 1,000h test passes:**
- Extrapolate to 10,000h using degradation rate
- Apply safety factor of 2×
- **Required:** <1% degradation at 1,000h → <10% at 10,000h

**If 1,000h test fails:**
- Identify failure mode (coating? electrode? PZT?)
- Redesign or reject supplier
- Do NOT proceed to production

### 6.3 Warranty Implications

**If supplier claims 10,000h but test shows 8,000h:**
- Negotiate warranty: replace at 8,000h (4 years at 6h/day)
- Or accept 8,000h with scheduled maintenance
- Or reject supplier (doesn't meet spec)

**C88 warranty strategy:**
- Customer warranty: 1 year (industry standard)
- Design target: 2 years (safety margin)
- Test requirement: 10,000h (3-5 years at 6-8h/day)

---

## 7. Cost and Timeline

### 7.1 Test Cost Estimate

**Phase 1 (500h, 3 suppliers):**
- Duration: 3 weeks
- Samples: 9 piezos (3 per supplier)
- Equipment: ¥20,000 (rental)
- Chemicals: ¥2,000
- Labor: ¥10,000
- **Total: ¥32,000** ($4,400)

**Phase 2 (1,000h, 2 suppliers):**
- Duration: 6 weeks
- Samples: 10 piezos (5 per supplier)
- Equipment: ¥30,000 (rental)
- Chemicals: ¥5,000
- Labor: ¥20,000
- **Total: ¥55,000** ($7,600)

**Phase 3 (2,000h, 1 supplier):**
- Duration: 12 weeks
- Samples: 6 piezos
- Equipment: ¥40,000 (rental)
- Chemicals: ¥10,000
- Labor: ¥30,000
- SEM analysis: ¥15,000
- **Total: ¥95,000** ($13,100)

**Grand total: ¥182,000** ($25,100) over 21 weeks

### 7.2 Timeline

```
Week 1-3:  Phase 1 (500h test, 3 suppliers)
Week 4:    Analyze results, select top 2
Week 5-10: Phase 2 (1,000h test, 2 suppliers)
Week 11:   Analyze results, select winner
Week 12-23: Phase 3 (2,000h test, 1 supplier)
Week 24:   Final analysis, supplier qualification
```

**Total: 24 weeks (6 months)**

### 7.3 Fast-Track Option

**If timeline is critical:**
- Skip Phase 1, go directly to Phase 2
- Test only SWT and Audiowell (most promising)
- Compress Phase 2 to 500h (higher acceleration)
- **Total: 12 weeks (3 months), ¥90,000** ($12,400)

**Risk:** May select wrong supplier, need to re-test later.

---

## 8. References

### 8.1 Standards and Guidelines

1. IEC 60068 — Environmental Testing Series  
2. ASTM B117 — Salt Spray (Fog) Testing  
3. MIL-HDBK-217F — Reliability Prediction of Electronic Equipment  
4. JEDEC JESD22-A108 — High Temperature Operating Life (HTOL)

### 8.2 Technical Papers

5. "Accelerated Life Testing of Piezoelectric Actuators" — IEEE Transactions on UFFC  
6. "Reliability of Piezoelectric Ceramics in Harsh Environments" — Journal of the American Ceramic Society  
7. "Lifetime Prediction of Ultrasonic Atomizers" — Ultrasonics Sonochemistry

### 8.3 Application Notes

8. Murata — "Piezoelectric Ceramic Reliability Guidelines"  
9. TDK — "Design Guidelines for Piezo Actuators"  
10. PI Ceramic — "Lifetime Testing of Piezo Components"

---

## 9. C88 Application Notes

### 9.1 Key Recommendations

1. **Start testing immediately** — 6-month timeline is critical path
2. **Use accelerated aging** — can't wait 14 months for real-time test
3. **Test all 3 chemicals** — HOCl, H₂O₂, PAA (not just water)
4. **Require test data from suppliers** — don't accept marketing claims
5. **Plan for failure** — design for easy piezo replacement

### 9.2 Questions for Saturday SWT Visit

1. Can you provide 3 samples for our 1,000h test?
2. Will you witness the test (or allow third-party witness)?
3. What is your warranty policy if piezos fail before 10,000h?
4. Do you have any test reports for HOCl/H₂O₂/PAA environments?
5. What is your recommended test protocol?

### 9.3 Risk Mitigation

**Risk: No supplier passes 1,000h test**
- Mitigation: Extend test to 2,000h with lower acceleration
- Or: Redesign system for easier piezo replacement (every 2 years)
- Or: Accept 8,000h lifetime with scheduled maintenance

**Risk: Test takes too long (6 months)**
- Mitigation: Fast-track option (3 months, higher risk)
- Or: Start production with SWT glass glaze (8,000h) while testing continues
- Or: Use Audiowell modules (claimed 10,000h) without full qualification

---

**End of Chapter 04 — Lifetime Testing Methods**

**Next:** Chapter 05 — Failure Modes



<!-- ===== Chapter 05: Failure Modes & Mechanisms / 失效模式与机理 ===== -->
# Piezoelectric Atomizer Failure Modes

**C88 Knowledge Base — Chapter 05**  
**Last Updated:** 2026-11-18  
**Purpose:** Understand how piezo atomizers fail, how to detect failures, and how to prevent them in C88

---

## 1. Failure Mode Overview

### 1.1 Common Failure Modes

**Top 7 failure modes for piezo atomizers in corrosive environments:**

1. **Cavitation erosion** — micro-bubble collapse damages surface
2. **Electrode corrosion** — silver electrode oxidizes/dissolves
3. **Coating delamination** — protective layer peels off
4. **Frequency drift** — resonance frequency shifts out of spec
5. **Cracking/fracture** — mechanical or thermal shock
6. **Depolarization** — loss of piezoelectric properties
7. **Contamination** — mineral deposits, biofilm

### 1.2 C88-Specific Risks

**High-risk failure modes for C88:**
- Cavitation erosion (continuous operation, high power)
- Electrode corrosion (HOCl/H₂O₂/PAA attack)
- Coating delamination (chemical + thermal cycling)
- Frequency drift (temperature + aging)

**Low-risk failure modes for C88:**
- Cracking (if handled properly)
- Depolarization (operating temp well below Curie)
- Contamination (filtered liquids, regular maintenance)

---

## 2. Cavitation Erosion

### 2.1 Mechanism

**How it happens:**
1. Ultrasonic vibration creates pressure waves in liquid
2. Low-pressure zones cause micro-bubbles to form (cavitation)
3. Bubbles collapse violently (implosion)
4. Shockwave hits piezo surface at high velocity (100-1000 m/s)
5. Repeated impacts erode surface material

**Damage characteristics:**
- Pitting (small craters on surface)
- Roughening (loss of smooth finish)
- Material loss (thinning of coating/electrode)
- Fatigue cracks (propagate from pits)

### 2.2 Factors Affecting Severity

**Increases cavitation:**
- Higher power (more intense pressure waves)
- Lower liquid level (less damping)
- Higher temperature (lower cavitation threshold)
- Dissolved gases (more bubble nucleation sites)
- Surface defects (scratch, pit = nucleation site)

**Decreases cavitation:**
- Smooth surface (fewer nucleation sites)
- Hard coating (resists erosion)
- Proper liquid level (adequate damping)
- Degassed liquid (fewer bubbles)

### 2.3 Detection Methods

**Visual inspection:**
- Pitting visible under microscope (50-200×)
- Surface roughening (loss of mirror finish)
- Discoloration (exposed electrode material)

**Performance monitoring:**
- Atomization rate decreases (erosion reduces efficiency)
- Drive current increases (more power needed for same output)
- Frequency shifts (mass loss changes resonance)

**Advanced methods:**
- SEM imaging (quantify pit size/density)
- Profilometry (measure surface roughness)
- Weight measurement (quantify material loss)

### 2.4 Prevention

**Design solutions:**
- Use hard coating (TiO₂, glass glaze)
- Smooth surface finish (Ra <0.1 µm)
- Adequate liquid level (always submerged)
- Avoid sharp corners (stress concentration)

**Operational solutions:**
- Don't run dry (always have liquid present)
- Filter liquid (remove particles that create pits)
- Control temperature (<50°C)
- Duty cycle operation (allow cooling)

**C88 recommendation:**
- Glass glaze or TiO₂ coating (hard, smooth)
- Liquid level sensor (prevent dry running)
- Temperature monitoring (shut down if >50°C)
- 0.5µm liquid filter (remove particles)

---

## 3. Electrode Corrosion

### 3.1 Mechanism

**How it happens:**
1. Silver electrode exposed to corrosive liquid (HOCl, H₂O₂, PAA)
2. Chemical reaction: Ag + Cl⁻ → AgCl (silver chloride)
3. AgCl is insoluble, forms white/grey deposit
4. Deposit increases contact resistance
5. Eventually electrode dissolves completely

**Galvanic corrosion:**
- Silver electrode + PZT ceramic + liquid = electrochemical cell
- Silver is anodic (corrodes preferentially)
- PZT contains lead (Pb²⁺) which is cathodic
- Current flows through liquid, accelerating corrosion

### 3.2 Factors Affecting Severity

**Increases corrosion:**
- High Cl⁻ concentration (HOCl dissociates to Cl⁻)
- Low pH (acidic environment accelerates corrosion)
- High temperature (faster chemical reactions)
- Exposed electrode (no protective coating)
- Dissimilar metals (galvanic couple)

**Decreases corrosion:**
- Protective coating (Parylene, glass glaze)
- Neutral pH (HOCl pH 5-6.5 is moderate)
- Low temperature (<40°C)
- Noble metal electrode (gold, platinum — expensive)

### 3.3 Detection Methods

**Visual inspection:**
- White/grey deposit on electrode (AgCl)
- Discoloration (brown/black = silver oxide)
- Pitting or roughening
- Exposed PZT (electrode completely corroded)

**Electrical measurements:**
- Increased contact resistance (measure with ohmmeter)
- Drive current increases (more power lost to resistance)
- Capacitance changes (electrode area reduced)

**Advanced methods:**
- EDS analysis (identify corrosion products)
- Cross-section SEM (see electrode thickness)
- Electrochemical impedance spectroscopy (EIS)

### 3.4 Prevention

**Design solutions:**
- Protective coating over electrode (Parylene, glass glaze)
- Noble metal electrode (gold, platinum — 10× cost)
- Encapsulated electrode (no liquid contact)
- Sacrificial anode (protect silver electrode)

**Operational solutions:**
- Regular rinsing (remove corrosive chemicals)
- pH control (keep near neutral)
- Temperature control (<40°C)
- Duty cycle (reduce exposure time)

**C88 recommendation:**
- Protective coating (Audiowell electrode film or Parylene)
- Regular rinse cycle (DI water after each use)
- Temperature monitoring (<50°C)
- Inspect electrodes every 6 months

---

## 4. Coating Delamination

### 4.1 Mechanism

**How it happens:**
1. Coating applied to PZT substrate (glass glaze, Parylene, TiO₂)
2. Thermal cycling causes stress (different CTE)
3. Chemical attack weakens interface
4. Coating lifts/peels from substrate
5. Exposed PZT/electrode corrodes rapidly

**Root causes:**
- Poor adhesion (surface contamination before coating)
- Thermal mismatch (CTE difference >5 ppm/°C)
- Chemical attack (liquid penetrates coating)
- Mechanical stress (vibration, handling)
- Manufacturing defect (pinhole, thin spot)

### 4.2 Factors Affecting Severity

**Increases delamination:**
- Large CTE mismatch (coating vs PZT)
- Thermal cycling (rapid temperature changes)
- Chemical attack (strong acids/bases)
- Poor surface preparation (contamination)
- Thin coating (<10 µm)

**Decreases delamination:**
- Matched CTE (coating ≈ PZT)
- Gradual temperature changes
- Chemical-resistant coating
- Proper surface preparation
- Adequate coating thickness (20-50 µm)

### 4.3 Detection Methods

**Visual inspection:**
- Blistering (bubbles under coating)
- Peeling/flaking (coating lifting)
- Discoloration (liquid under coating)
- Exposed substrate (coating completely gone)

**Adhesion tests:**
- Tape test (ASTM D3359) — apply tape, pull off, check coating
- Scratch test — scratch with stylus, check for delamination
- Pull-off test — measure adhesion strength (MPa)

**Advanced methods:**
- Ultrasonic C-scan (detect subsurface delamination)
- Thermography (IR camera detects delamination)
- Dye penetrant (dye seeps into delaminated areas)

### 4.4 Prevention

**Design solutions:**
- Matched CTE (coating ≈ PZT)
- Adhesion promoter (silane coupling agent)
- Multi-layer coating (gradual CTE transition)
- Adequate thickness (20-50 µm)

**Manufacturing solutions:**
- Proper surface preparation (clean, etch, prime)
- Controlled application (temperature, humidity)
- Curing/firing (proper temperature profile)
- Quality inspection (adhesion test every batch)

**C88 recommendation:**
- Require adhesion test from supplier (ASTM D3359, Class 4B minimum)
- Inspect coatings upon receipt (visual + tape test)
- Avoid thermal shock (gradual temperature changes)
- Replace if delamination observed (don't continue operation)

---

## 5. Frequency Drift

### 5.1 Mechanism

**How it happens:**
1. Piezo resonance frequency determined by material properties + geometry
2. Aging changes material properties (domain wall pinning)
3. Temperature changes elastic constants
4. Coating erosion changes mass/thickness
5. Mechanical stress changes dimensions
6. Frequency shifts away from design value

**Types of drift:**
- **Aging drift** — gradual, irreversible (hours to months)
- **Temperature drift** — reversible, follows temperature (°C)
- **Load drift** — reversible, follows liquid level/pressure
- **Stress drift** — irreversible, from mechanical shock

### 5.2 Factors Affecting Severity

**Increases drift:**
- High temperature (accelerates aging)
- High drive power (self-heating)
- Thermal cycling (repeated expansion/contraction)
- Mechanical shock (vibration, handling)
- Coating erosion (mass loss)

**Decreases drift:**
- Stable temperature (controlled environment)
- Low drive power (less self-heating)
- Gradual temperature changes
- Careful handling (no shock)
- Stable coating (no erosion)

### 5.3 Detection Methods

**Direct measurement:**
- Impedance analyzer (measure resonance frequency)
- Network analyzer (S11 parameter, resonance peak)
- Frequency counter (monitor driver output)

**Indirect measurement:**
- Atomization rate decreases (frequency off-optimum)
- Drive current increases (impedance mismatch)
- Temperature rises (inefficient operation)

**Continuous monitoring:**
- PLL frequency tracking (driver locks to resonance)
- Phase detection (monitor impedance phase angle)
- Current monitoring (detect impedance changes)

### 5.4 Prevention

**Design solutions:**
- Temperature compensation (TCXO in driver)
- Frequency tracking (PLL locks to resonance)
- Low-stress mounting (avoid clamping forces)
- Stable coating (resists erosion)

**Operational solutions:**
- Temperature control (<50°C)
- Gradual warm-up (avoid thermal shock)
- Regular calibration (check frequency monthly)
- Replace if drift >2% (out of spec)

**C88 recommendation:**
- TCXO driver (±0.5 ppm/°C stability)
- Frequency tracking (PLL or phase detection)
- Temperature monitoring (shut down if >50°C)
- Monthly frequency check (impedance analyzer)
- Replace if drift >2% from initial

---

## 6. Cracking and Fracture

### 6.1 Mechanism

**How it happens:**
1. PZT is brittle ceramic (low fracture toughness)
2. Mechanical shock creates crack
3. Thermal shock creates crack (rapid temperature change)
4. Crack propagates under vibration (fatigue)
5. Piezo fractures completely

**Types of cracks:**
- **Radial cracks** — from center to edge (mechanical shock)
- **Concentric cracks** — ring-shaped (thermal stress)
- **Edge chips** — small pieces broken off (handling damage)
- **Complete fracture** — piezo breaks into pieces (catastrophic)

### 6.2 Factors Affecting Severity

**Increases cracking:**
- Mechanical shock (dropping, impact)
- Thermal shock (rapid temperature change)
- High vibration amplitude (high drive power)
- Sharp corners (stress concentration)
- Surface flaws (scratch, pit)

**Decreases cracking:**
- Careful handling (no dropping)
- Gradual temperature changes
- Moderate drive power
- Rounded corners (reduce stress)
- Smooth surface (no flaws)

### 6.3 Detection Methods

**Visual inspection:**
- Visible cracks (naked eye or microscope)
- Edge chips (missing pieces)
- Complete fracture (broken into pieces)

**Electrical tests:**
- Open circuit (complete fracture)
- Capacitance drop (crack reduces electrode area)
- Insulation resistance drop (crack allows current leakage)

**Advanced methods:**
- Ultrasonic testing (detect subsurface cracks)
- Dye penetrant (dye seeps into cracks)
- X-ray inspection (see internal cracks)

### 6.4 Prevention

**Design solutions:**
- Rounded corners (reduce stress concentration)
- Compliant mounting (rubber gasket, silicone potting)
- Protective housing (metal can, ceramic shell)
- Redundant design (multiple piezos, fail-safe)

**Manufacturing solutions:**
- Careful handling (tweezers, anti-static)
- Proper packaging (foam, anti-static bags)
- Quality inspection (visual + electrical test)
- Burn-in screening (detect early failures)

**C88 recommendation:**
- Compliant mounting (silicone gasket)
- Protective housing (316L can)
- Careful handling procedures (training, packaging)
- Incoming inspection (visual + electrical test)
- Replace immediately if crack detected

---

## 7. Depolarization

### 7.1 Mechanism

**How it happens:**
1. PZT is polarized during manufacturing (aligned domains)
2. High temperature randomizes domains (above Curie temp)
3. High electric field reverses polarization (overvoltage)
4. Mechanical stress disrupts domains (shock, vibration)
5. Piezo loses piezoelectric properties

**Types of depolarization:**
- **Thermal depolarization** — temperature > Curie point
- **Electrical depolarization** — voltage > coercive field
- **Mechanical depolarization** — stress > coercive stress

### 7.2 Factors Affecting Severity

**Increases depolarization:**
- High temperature (>200°C for PZT-4)
- High electric field (>2 kV/mm)
- High mechanical stress (>100 MPa)
- AC electric field (cycling causes fatigue)

**Decreases depolarization:**
- Low temperature (<100°C)
- Moderate electric field (<1 kV/mm)
- Low mechanical stress (<50 MPa)
- DC bias (stabilizes domains)

### 7.3 Detection Methods

**Electrical measurements:**
- Capacitance drop (depolarized piezo has lower C)
- d₃₃ measurement (piezoelectric coefficient)
- Impedance spectrum (resonance peak disappears)

**Performance measurements:**
- Atomization rate drops (no vibration)
- Drive current drops (no piezo effect)
- Temperature rises (energy not converted to vibration)

### 7.4 Prevention

**Design solutions:**
- High Curie temperature PZT (PZT-4: 328°C)
- Voltage limiting (protect against overvoltage)
- Stress relief (compliant mounting)
- DC bias (stabilize polarization)

**Operational solutions:**
- Temperature control (<100°C)
- Voltage control (<1 kV/mm)
- Avoid mechanical shock
- Regular testing (check capacitance monthly)

**C88 recommendation:**
- PZT-4 or PZT-8 (high Curie temp)
- Operating temp <50°C (well below Curie)
- Drive voltage <100Vpp (<0.1 kV/mm for 1mm thick)
- Compliant mounting (no mechanical stress)
- Depolarization unlikely in C88 application

---

## 8. Contamination

### 8.1 Mechanism

**How it happens:**
1. Mineral deposits from hard water (Ca²⁺, Mg²⁺)
2. Biofilm from bacteria/fungi growth
3. Dust/particles from air
4. Chemical residues from cleaning agents
5. Contaminants block atomization, reduce efficiency

### 8.2 Detection Methods

**Visual inspection:**
- White/green deposits (minerals, biofilm)
- Cloudy liquid (suspended particles)
- Reduced atomization rate

**Performance monitoring:**
- Atomization rate decreases
- Drive current increases
- Frequency shifts

### 8.3 Prevention

**Design solutions:**
- Liquid filtration (0.5µm filter)
- Regular rinse cycle (DI water)
- Sealed system (prevent dust ingress)
- Anti-bacterial coating (silver ions)

**Operational solutions:**
- Use DI or filtered water
- Regular cleaning (weekly rinse)
- Replace liquid regularly (prevent stagnation)
- Inspect and clean monthly

**C88 recommendation:**
- 0.5µm liquid filter (remove particles)
- DI water rinse after each use
- Weekly cleaning (vinegar solution for minerals)
- Monthly inspection (visual + performance test)

---

## 9. Failure Mode Summary

### 9.1 C88 Risk Matrix

| Failure Mode | Probability | Severity | Risk Level | Mitigation |
|--------------|-------------|----------|------------|------------|
| Cavitation erosion | Medium | High | **High** | Hard coating, liquid level sensor |
| Electrode corrosion | High | High | **Critical** | Protective coating, regular rinse |
| Coating delamination | Medium | High | **High** | Adhesion test, thermal control |
| Frequency drift | Medium | Medium | **Medium** | TCXO driver, temperature control |
| Cracking/fracture | Low | High | **Medium** | Careful handling, compliant mount |
| Depolarization | Low | High | **Low** | Temp <50°C, voltage <100Vpp |
| Contamination | High | Low | **Medium** | Filtration, regular cleaning |

### 9.2 Critical Failure Modes for C88

**Must address:**
1. **Electrode corrosion** — HOCl/H₂O₂/PAA attack (use protective coating)
2. **Cavitation erosion** — continuous operation (use hard coating)
3. **Coating delamination** — thermal cycling (use matched CTE)

**Monitor but not critical:**
4. Frequency drift — TCXO driver compensates
5. Contamination — regular cleaning prevents
6. Cracking — careful handling prevents

**Unlikely in C88:**
7. Depolarization — operating conditions well within limits

### 9.3 Maintenance Strategy

**Daily:**
- Visual inspection (check for damage, contamination)
- Performance check (atomization rate, frequency)

**Weekly:**
- Cleaning (DI water rinse, vinegar for minerals)
- Electrical test (frequency, capacitance, current)

**Monthly:**
- Detailed inspection (microscope for cracks, pitting)
- Adhesion test (tape test for coating)
- Calibration (impedance analyzer)

**Annually:**
- Full replacement (after 8,000-10,000h)
- Or inspection + testing (if lifetime extended)

---

## 10. References

### 10.1 Technical Papers

1. "Cavitation Erosion of Piezoelectric Ceramics" — Wear Journal  
2. "Corrosion of Silver Electrodes in Chloride Solutions" — Corrosion Science  
3. "Delamination Mechanisms in Thin Film Coatings" — Surface and Coatings Technology  
4. "Frequency Stability of Piezoelectric Resonators" — IEEE Transactions on UFFC  
5. "Fracture Mechanics of Piezoelectric Ceramics" — Journal of the American Ceramic Society

### 10.2 Standards

6. ASTM D3359 — Standard Test Methods for Measuring Adhesion by Tape Test  
7. ASTM G85 — Standard Practice for Modified Salt Spray (Fog) Testing  
8. IEC 60068 — Environmental Testing Series  
9. MIL-STD-810 — Environmental Engineering Considerations

### 10.3 Application Notes

10. Murata — "Reliability Guidelines for Piezoelectric Ceramics"  
11. TDK — "Design Guidelines for Piezo Actuators"  
12. PI Ceramic — "Lifetime and Failure Modes of Piezo Components"

---

## 11. C88 Application Notes

### 11.1 Key Takeaways

1. **Electrode corrosion is the #1 risk** — must use protective coating
2. **Cavitation erosion is #2 risk** — hard coating extends life
3. **Coating delamination is #3 risk** — adhesion test critical
4. **Frequency drift manageable** — TCXO driver compensates
5. **Other failure modes unlikely** — proper design/operation prevents

### 11.2 Design Recommendations

**Coating selection:**
- Best: Audiowell electrode film (Parylene?)
- Alternative: TiO₂ sputtering (if qualified)
- Acceptable: Glass glaze (8,000h lifetime)

**Driver design:**
- TCXO oscillator (±0.5 ppm/°C)
- Frequency tracking (PLL or phase detection)
- Current limiting (protect against overload)
- Temperature monitoring (shut down if >50°C)

**Mechanical design:**
- Compliant mounting (silicone gasket)
- Protective housing (316L can)
- Liquid level sensor (prevent dry running)
- 0.5µm filter (remove particles)

### 11.3 Maintenance Recommendations

**Daily:** Visual inspection, performance check  
**Weekly:** Cleaning, electrical test  
**Monthly:** Detailed inspection, adhesion test  
**Annually:** Full replacement or extended testing

### 11.4 Questions for Saturday SWT Visit

1. What is the most common failure mode you've seen?
2. How do you test for coating adhesion?
3. What is your warranty policy for coating delamination?
4. Can you provide failure analysis reports?
5. What is your recommended maintenance schedule?

---

**End of Chapter 05 — Failure Modes**

**Knowledge Base Complete:**
- Chapter 01: Piezoelectric Atomization Physics ✓
- Chapter 02: Anti-Corrosion Coating Technologies ✓
- Chapter 03: Driver Circuit Design ✓
- Chapter 04: Lifetime Testing Methods ✓
- Chapter 05: Failure Modes ✓

**Total:** 5 chapters, ~75KB of technical knowledge



<!-- ===== Chapter 06: Chinese Technical References / 中国压电雾化技术资料 ===== -->
# 中国压电雾化技术参考资料

**C88 Knowledge Base — 中国技术补充**  
**Created:** 2026-11-18  
**Purpose:** 补充中国压电雾化技术资料，为周六SWT拜访做准备

---

## 1. 百度百科 — 雾化片

**来源:** https://baike.baidu.com/item/%E9%9B%BE%E5%8C%96%E7%89%87/1629739

### 关键技术参数

| 参数 | 数值 | 备注 |
|------|------|------|
| 常见直径 | 13.8mm, 16mm, 20mm, 25mm | 25mm为最大 |
| 工作频率 | 1.7MHz, 2.4MHz, 3.0MHz | 高频可达3MHz |
| 驱动电压 | 24-48V (陶瓷雾化片) | 微孔雾化片3-5V |
| 使用寿命 | ≥5000小时 | 部分产品≥8000小时 |
| 雾化颗粒 | 3-5µm (常规) | 先进技术可达0.5µm |
| 谐振阻抗 | ≤2.0Ω 或 ≤180Ω | 取决于型号 |
| 静态电容 | 1100-2000pF ±15% | 典型值 |
| 最大工作电压 | ≤60V | |
| 最大工作温度 | ≤85°C | |

### 防腐技术

**镍电极 vs 银电极:**
- 银电极：传统，易腐蚀
- 镍电极：90年代发明，防腐性和耐久性远超银电极
- 来源：东方金荣Siansonic于90年代发明镍电极压电陶瓷雾化换能器

**特殊防护涂层:**
- 镍电极可实现耐化学和电腐蚀
- 谐振频率可达3MHz，产生平均粒径仅1µm的超细雾滴

### 亚微米雾化

**先进技术:**
- 提高谐振频率 + 特殊结构设计
- 可实现平均粒径0.5µm的亚微米级气溶胶雾化
- 东方金荣已实现此技术

**C88相关:**
- C88目标：0.5-5µm液滴
- 1.7MHz → 2.3µm (常规)
- 需要特殊设计才能达到亚微米级

---

## 2. 东方金荣Siansonic — 超声波雾化技术

**来源:** https://zhuanlan.zhihu.com/p/591727437  
**作者:** 东方金荣公司 (Siansonic)  
**发布日期:** 2022-12-14

### 三种超声波雾化方式

#### 方式1：单晶片压电陶瓷式

**原理:**
- 压电陶瓷换能器在液体中振动发射超声波
- 超声波传递到液-气交界面
- 声阻抗差异导致能量聚集，撕裂液体成微滴

**历史:**
- 60-70年代：医用雾化吸入（最早应用）
- 随后日本用于环境加湿
- 90年代：东方金荣发明镍电极压电陶瓷，防腐性远超银电极

**技术参数:**
- 频率：1-3MHz
- 液滴粒径：3-5µm（常规）
- 功率：20W以上 → 300ml/h雾化量
- 可雾化液体最大粘度：1.2cps（接近水）
- 必须浸入液体中工作
- 需要一定液位高度和成雾高度

**优缺点:**
- ✅ 结构简单，成本低
- ✅ 超声频率高，雾化颗粒小
- ❌ 能量转化效率低（300ml/h需20W+）
- ❌ 必须浸入液体
- ❌ 雾化方向受限（不能自上而下）
- ❌ 只能雾化低粘度液体（≤1.2cps）

**C88适用性:**
- C88使用单晶片压电陶瓷式
- 1.7MHz频率，2.3µm液滴
- 12片压电片，总功率240-480W
- 目标输出：600-1200g/h

#### 方式2：微孔网片式

**原理:**
- 环形压电陶瓷带动微孔网片轴向振动
- 液体从微孔（5-10µm）挤出形成雾化
- 雾化粒径与频率无关，仅与孔径有关

**技术参数:**
- 驱动电压：3-5V
- 功率：1-2W
- 最大雾化量：<10ml/h
- 可雾化液体最大粘度：1-2cps

**优缺点:**
- ✅ 效率高（3-5V，1-2W）
- ✅ 喷雾方向自由
- ✅ 不需要累积液体
- ❌ 雾化量很低（<10ml/h）
- ❌ 微孔易堵塞
- ❌ 只能雾化少量低粘度液体

**C88适用性:**
- ❌ 不适合C88（雾化量太低）
- 适用于便携式微量雾化（香薰器、手持雾化吸入器）

#### 方式3：朗之万换能器式

**原理:**
- 朗之万式超声换能器 + 变幅杆
- 液体输送到变幅杆最大振幅点前端
- 超声振动撕裂液体成雾

**技术参数:**
- 频率：20-200kHz（低频）
- 最小雾化颗粒：≥10µm
- 可雾化液体最大粘度：30-50cps（水的30-50倍）
- 功率：1W → 1200ml/h（惊人！）
- 任意方向喷雾
- "送达即雾化"

**优缺点:**
- ✅ 雾化能力极强（粘度30-50cps）
- ✅ 任意方向喷雾
- ✅ 精确即时雾化（纳升/秒级）
- ❌ 雾化粒径大（≥10µm）
- ❌ 频率不能很高（20-200kHz）

**C88适用性:**
- ❌ 不适合C88（液滴太大，≥10µm）
- 适用于薄膜涂层、喷雾热解等工业应用

### 三种方式对比表

| 特性 | 单晶片压电陶瓷 | 微孔网片 | 朗之万换能器 |
|------|---------------|---------|-------------|
| 频率 | 1-3MHz | 110-170kHz | 20-200kHz |
| 液滴粒径 | 3-5µm | 5-10µm（孔径） | ≥10µm |
| 驱动电压 | 24-48V | 3-5V | 高电压 |
| 功率效率 | 20W→300ml/h | 1-2W→<10ml/h | 1W→1200ml/h |
| 液体粘度 | ≤1.2cps | ≤1-2cps | 30-50cps |
| 喷雾方向 | 受限（需浸入） | 自由 | 任意 |
| 应用 | 加湿、医疗 | 便携式微量 | 薄膜涂层、工业 |

### 东方金荣技术能力

**公司背景:**
- 近40年超声波雾化技术经验
- 掌握全部三种雾化方式
- 探索前沿技术：声表面波雾化、高强度聚焦超声雾化、非接触声场雾化

**亚微米雾化:**
- 已实现平均粒径0.5µm的亚微米级气溶胶雾化
- 用于喷雾热解法超细粉体制备
- 应用：先进材料制造领域

---

## 3. 电子工程专辑 — 雾化片驱动电路

**来源:** https://eet-china.com/mp/a130957.html  
**作者:** ittbank  
**发布日期:** 2022-05-11

### 驱动电路类型

#### 方式1：分立元件驱动

**电路组成:**
- 电源电路（变压器降压 + 桥式整流）
- 振荡器电路
- 水位控制电路

**工作原理:**
1. 220V AC → 变压器降压 → 桥式整流 → DC
2. 水位电极检测有水 → V1导通 → V2导通
3. 振荡器工作 → 雾化片高频振动 → 雾化

**关键元器件:**
- R1-R7: 1/4W碳膜/金属膜电阻
- C1: 50V+铝电解电容
- VD1-VD5: 1N4007整流二极管
- VS1-VS2: 4.7V稳压二极管
- V1: S9015或S8550 (PNP)
- V2: 2SC1815或S8050 (NPN)
- V3: BU406或MJEL3005 (NPN高反压)
- T: 30W, 二次电压36V变压器
- B: 陶瓷雾化片（1.7MHz或2.4MHz）

**特点:**
- 简单，成本低
- 适合家用加湿器
- 频率固定（1.7MHz或2.4MHz）

#### 方式2：专用IC驱动（SJ303）

**SJ303功能:**
- 高频PWM发生器
- DC-DC升压控制器
- MCU内核

**特点:**
- 集成度高
- 需要外部MCU控制
- 适用于微孔雾化片

#### 方式3：专用IC驱动（TTP320）

**TTP320两个版本:**
- TTP320-AO8: 需要搭配MCU
- TTP320-BO8: 独立工作，不需要MCU

**特点:**
- 专用于微孔雾化片
- 简化设计
- 低成本

### 关键设计要点

**频率匹配:**
- 驱动板振荡电路针对固定频率调校
- 2.4MHz雾化片接入1.7MHz驱动板 → 阻抗极高，几乎不起振
- 大部分电能转化为热能 → 迅速烧毁
- **必须匹配频率！**

**C88设计建议:**
- 使用分立元件驱动（1.7MHz或2.4MHz）
- 或专用IC驱动（如SJ303）
- 确保频率匹配
- 考虑温度补偿（TCXO）

---

## 4. 德瓷技术 Dc-piezo — 可靠性测试

**来源:** https://dc-piezo.com/reliability-test  
**公司:** 广东德瓷技术有限公司

### 可靠性测试体系

#### 1. 环境适应性测试

**温度循环测试:**
- 范围：-40°C 至 150°C
- 目的：验证材料热膨胀系数匹配性、内部结构稳定性

**湿热老化测试:**
- 条件：85°C / 85%RH
- 时长：1000小时
- 评估：陶瓷绝缘性能、电极抗腐蚀能力

**盐雾测试:**
- 模拟：海洋或工业腐蚀环境
- 检验：器件封装工艺防护效能

#### 2. 机械可靠性测试

**振动与冲击测试:**
- 随机振动：5-2000Hz
- 多轴冲击：50g加速度
- 确保：结构无裂纹、分层

**疲劳寿命测试:**
- 循环次数：百万次以上
- 监测：位移输出衰减率
- 确保：驱动精度长期稳定

#### 3. 电学性能稳定性测试

**高压击穿测试:**
- 施加电压：额定电压2-3倍
- 验证：介电强度、绝缘层可靠性

**阻抗-频率特性分析:**
- 仪器：精密LCR测试仪
- 检测：谐振频率、机电耦合系数一致性
- 确保：信号响应精准度

**负载老化测试:**
- 条件：额定电压下连续工作500小时
- 监测：介电损耗、压电常数(d33)变化率

#### 4. 寿命预测与失效分析

**加速寿命试验 (ALT):**
- 模型：阿伦尼乌斯模型 (Arrhenius)
- 方法：高温、高电压加速老化
- 推算：正常使用条件下的MTBF（平均无故障时间）

**微观结构分析:**
- SEM扫描电镜：分析晶界、孔隙等微观缺陷
- XRD衍射仪：分析晶体结构
- 优化：烧结工艺

### 质量管控体系

**DFR (可靠性设计):**
- 材料配方阶段导入可靠性仿真模型
- 电极结构设计阶段考虑可靠性

**SPC (统计过程控制):**
- 实时监控：烧结温度、极化电压等关键工艺参数
- 确保：批次一致性

**100%全检 + 抽样复测:**
- 出厂前：关键性能参数全检
- 定期：抽样进行破坏性测试

### C88相关测试建议

**参考德瓷测试标准:**

1. **湿热老化测试**
   - 条件：85°C / 85%RH, 1000小时
   - 评估：HOCl环境下的绝缘性能、电极腐蚀

2. **疲劳寿命测试**
   - 循环次数：百万次以上（模拟10,000小时）
   - 监测：雾化率衰减、频率漂移

3. **加速寿命试验 (ALT)**
   - 基于阿伦尼乌斯模型
   - 高温（60°C）+ 高浓度HOCl（2000ppm）
   - 推算10,000小时寿命

4. **微观结构分析**
   - SEM检查涂层完整性
   - XRD分析晶体结构变化

---

## 5. 中国压电雾化技术总结

### 技术优势

1. **镍电极技术** (90年代发明)
   - 防腐性远超传统银电极
   - 东方金荣Siansonic首创
   - 适用于HOCl/H₂O₂/PAA环境

2. **亚微米雾化技术**
   - 平均粒径0.5µm
   - 提高频率 + 特殊结构设计
   - 东方金荣已实现

3. **完整产业链**
   - 材料（PZT陶瓷）→ 元件（雾化片）→ 模组（驱动板）→ 系统
   - 奥迪威、德瓷、东方金荣等完整解决方案

4. **可靠性测试体系**
   - 德瓷建立完整测试体系
   - 环境、机械、电学、寿命全覆盖
   - 阿伦尼乌斯模型加速老化

### 与C88相关的关键信息

**雾化方式选择:**
- C88使用单晶片压电陶瓷式（1.7MHz）
- 液滴粒径2.3µm（常规）
- 需要12片（6片/模组 × 2模组）

**防腐技术:**
- 镍电极（优于银电极）
- 玻璃釉涂层（SWT, J&H）
- 电极保护膜（奥迪威）
- 钛膜（Xizhe）

**驱动电路:**
- 分立元件驱动（简单、低成本）
- 专用IC驱动（SJ303, TTP320）
- 必须频率匹配（1.7MHz或2.4MHz）

**寿命测试:**
- 参考德瓷测试标准
- 湿热老化：85°C/85%RH, 1000小时
- 加速寿命：阿伦尼乌斯模型
- 目标：10,000小时

### 周六SWT拜访重点问题

**基于中国技术资料:**

1. **镍电极 vs 银电极:**
   - 你们使用镍电极还是银电极？
   - 镍电极的防腐性能数据？

2. **玻璃釉涂层:**
   - 玻璃釉成分？（SiO₂/B₂O₃/Al₂O₃比例？）
   - 涂层厚度？（20µm? 50µm?）
   - 附着力测试方法？

3. **亚微米雾化:**
   - 能否实现0.5µm亚微米雾化？
   - 需要特殊结构设计吗？

4. **寿命测试:**
   - 是否做过85°C/85%RH湿热老化测试？
   - 加速寿命试验方法？（阿伦尼乌斯模型？）
   - 8,000小时寿命的测试条件？

5. **驱动电路:**
   - 使用分立元件还是专用IC？
   - 频率匹配精度？（±0.05MHz?）
   - 温度补偿方案？

---

## 6. 参考文献

### 中文资料

1. 百度百科 — 雾化片  
   URL: https://baike.baidu.com/item/%E9%9B%BE%E5%8C%96%E7%89%87/1629739  
   **关键内容:** 技术参数、防腐技术、亚微米雾化

2. 东方金荣Siansonic — 什么是超声波雾化？(知乎)  
   URL: https://zhuanlan.zhihu.com/p/591727437  
   **关键内容:** 三种雾化方式对比、镍电极发明历史、亚微米雾化技术

3. 电子工程专辑 — 探秘：雾化片的驱动电路  
   URL: https://eet-china.com/mp/a130957.html  
   **关键内容:** 分立元件驱动、专用IC驱动、频率匹配要点

4. 广东德瓷技术有限公司 — 可靠性测试  
   URL: https://dc-piezo.com/reliability-test  
   **关键内容:** 环境测试、机械测试、电学测试、加速寿命试验

5. 奥迪威官网 — 超声波雾化片  
   URL: https://audiowell.net/atomizingtransducer.html  
   **关键内容:** 电极保护膜技术、自主开发压电材料

### 英文资料（已在知识库其他章节）

6. Lang, R.J. (1962). "Ultrasonic atomization of liquids". JASA 34(1), 6-8.
7. Wikipedia: Ultrasonic atomization
8. Parylene Coating Properties — HZO Inc.

---

**End of Chinese Technical References**

**知识库更新完成:**
- 01-压电物理 (14KB) ✓
- 02-防腐涂层 (19KB) ✓
- 03-驱动电路 (15KB) ✓
- 04-寿命测试 (14KB) ✓
- 05-失效模式 (20KB) ✓
- **中国技术补充 (新增)** ✓
- README索引 (9KB) ✓

**总计：~100KB技术知识，含中英文资料**



<!-- ===== Chapter 07: Chinese Academic Papers & Patents / 中国学术论文与专利 ===== -->
# 中国学术论文与专利资料

**C88 Knowledge Base — 中国学术研究**  
**Created:** 2026-11-18  
**Purpose:** 中国大学和研究机构的压电雾化技术学术论文，为C88提供理论支撑

---

## 1. 清华大学 — 高性能无铅压电陶瓷

### 1.1 研究团队

**王轲研究员 + 李敬锋教授**  
清华大学材料学院

### 1.2 研究成果

**论文1: 无铅压电陶瓷高压电响应机制**
- **标题:** "Deciphering the phase transition-induced ultrahigh piezoresponse in (K,Na)NbO3-based piezoceramics"
- **期刊:** Nature Communications (2022)
- **DOI:** https://doi.org/10.1038/s41467-022-31158-x
- **关键发现:**
  - KNN基无铅压电陶瓷的压电常数(d33)已可媲美含铅压电陶瓷
  - 在超声雾化、水声换能等领域逐步替代传统PZT
  - 揭示了"场致相变"新机制：相变诱导的体积变化贡献了1250 pm/V的等效压电常数
  - 使用原位同步辐射、高分辨透射电镜等先进表征手段

**论文2: 氧空位缺陷设计**
- **标题:** "Isolated-Oxygen-Vacancy Hardening in Lead-Free Piezoelectrics"
- **期刊:** Advanced Materials (2022)
- **DOI:** https://doi.org/10.1002/adma.202202558
- **关键发现:**
  - 通过调节氧空位浓度，在d33几乎不变的基础上，Qm提升60%以上
  - 解决了Qm与d33此消彼长的制约关系
  - 氧空位引起晶格畸变与应力失配，抑制畴壁运动
  - 大功率工况下服役稳定性明显优于同类材料

**C88应用价值:**
- 无铅压电陶瓷可替代PZT（环保要求）
- 高Qm材料适合大功率连续工作（C88需要12片同时工作）
- 氧空位设计可提升寿命

**来源:** https://www.tsinghua.edu.cn/info/1175/96754.htm

---

## 2. 上海交通大学 — 无铅压电陶瓷驱动与传感

### 2.1 研究团队

**郭益平教授课题组**  
上海交通大学材料科学与工程学院  
金属基复合材料国家重点实验室

### 2.2 研究成果

**论文1: 缺陷偶极子诱导应变**
- **标题:** "Defect Dipole-Induced Fatigue Strain Recoverability in Lead-Free Piezoelectric Ceramics"
- **期刊:** Advanced Functional Materials (2023)
- **DOI:** https://doi.org/10.1002/adfm.202306416
- **关键发现:**
  - 在KNN基无铅压电陶瓷中获得3080 pm/V的超高逆压电系数
  - 发现独特的应变疲劳恢复特性
  - 验证了缺陷偶极子与单极应变性能的密切关联
  - 为设计高可靠性、大位移行程的无铅驱动材料提供新思路

**论文2: 锰掺杂KNN基压电陶瓷**
- **标题:** "Superior Electromechanical Compatibility in Lead-free Piezoceramics with Mobile Transition-Metal Defects"
- **期刊:** ACS Applied Materials & Interfaces (2023)
- **DOI:** http://pubs.acs.org/doi/abs/10.1021/acsami.3c12068
- **关键发现:**
  - 退火热处理后，d33提升20%，Qm提升80%
  - 综合性能可与商用PZT-5A陶瓷相媲美
  - **雾化片性能已超过同类PZT雾化片水平**
  - 锰离子在KNN基陶瓷中可同时占据A位与B位
  - 低价态Mn存在从B位向A位迁移的趋势

**论文3: 柔性压电传感器**
- **标题:** "Modulation of piezoelectricity and mechanical strength in piezoelectric composites based on N0.5B0.51T-BNT nanocubes towards human-machine interfaces"
- **期刊:** Nano Energy (2023)
- **DOI:** https://doi.org/10.1016/j.nanoen.2023.109044
- **关键发现:**
  - 钛酸铋钠基纳米单晶 + PDMS 0-3复合
  - 输出电压33.4V，电流1.8μA
  - 灵敏度1.94 V/N，响应时间27ms
  - 耐久性6000次循环

**C88应用价值:**
- 无铅压电陶瓷雾化片性能已超PZT
- 锰掺杂可大幅提升Qm（适合大功率）
- 缺陷工程可提升可靠性

**来源:** https://news.sjtu.edu.cn/jdzh/20231103/190096.html

---

## 3. 中南大学 — 无铅压电陶瓷材料与器件

### 3.1 研究团队

**周学凡特聘副研究员**  
中南大学粉末冶金研究院  
"新型电子功能材料与器件"团队（张斗教授领衔）

### 3.2 研究成果

**研究方向:**
- 无铅压电陶瓷材料与器件
- 铁电材料微纳结构合成与器件

**关键成果:**
- 开发了大压电系数、高场致应变、高能量存储和超高温无铅压电陶瓷新体系
- 揭示陶瓷材料中优异电学性能的微观结构起源
- 建立压电陶瓷相组成、畴结构、局域结构、缺陷结构等与宏观电学性能的相关关系
- 研制高性能无铅压电驱动器、**超声雾化片**、介电电容器和高温传感器

**发表论文:**
- 以第一或通讯作者在Nature Communications, Advanced Materials, Advanced Functional Materials, Progress in Materials Science等权威期刊发表论文30余篇

**团队主页:** http://www.csupiezogroup.com/

**C88应用价值:**
- 专门研究超声雾化片
- 无铅压电陶瓷替代PZT
- 可联系合作开发C88专用雾化片

**来源:** https://faculty.csu.edu.cn/zhouxuefan/zh_CN/index/208462/list/index.htm

---

## 4. 中国地质大学 — 压电式超声雾化降尘系统

### 4.1 研究团队

**韩光超教授**  
中国地质大学（武汉）机械学院

### 4.2 研究成果

**论文: 压电式超声雾化降尘系统设计及其试验研究**
- **期刊:** 机械科学与技术 (2024)
- **卷期:** 43(2): 258-264
- **研究内容:**
  - 压电式超声雾化降尘系统设计
  - 试验研究与性能评估
  - 应用于工业降尘领域

**其他相关论文:**
- 华中科技大学学报(自然科学版), 2014, 42(5): 52~56
- 西安交通大学学报, 2016, 50(9): 119-124

**C88应用价值:**
- 超声雾化在工业环境应用经验
- 可参考降尘系统的雾化效率优化

**来源:** https://jidian.cug.edu.cn/info/3061/30231.htm

---

## 5. 南京航空航天大学 — 压电雾化装置综述

### 5.1 论文信息

**标题:** 压电雾化装置的研究进展  
**期刊:** 南京航空航天大学学报 (2020)  
**URL:** https://tnuaa.nuaa.edu.cn/njhkhten/article/pdf/202001005

### 5.2 研究内容

**综述内容:**
1. **理论研究:**
   - 压电雾化机理
   - 液滴尺寸预测模型
   - 雾化效率分析

2. **应用研究:**
   - 喷墨打印
   - 喷雾冷却
   - 吸入治疗（医疗雾化）

3. **功能性分类:**
   - 单晶片压电陶瓷式
   - 微孔网片式
   - 朗之万换能器式

**关键结论:**
- 压电雾化装置具有结构简单、易携带、耗能低、生产成本低及雾化效果好等优点
- 广泛应用于吸入治疗、喷墨打印和喷雾冷却等领域

**C88应用价值:**
- 全面的压电雾化技术综述
- 可参考各类雾化方式的优缺点
- C88使用单晶片压电陶瓷式

---

## 6. 物理学报 — 铁电陶瓷失效分析

### 6.1 论文信息

**标题:** 铁电陶瓷脉冲耐压失效分布及耐压可靠性  
**期刊:** 物理学报 (2024)  
**DOI:** https://doi.org/10.7498/aps.73.20231354  
**URL:** https://wulixb.iphy.ac.cn/article/doi/10.7498/aps.73.20231354

### 6.2 研究内容

**研究对象:** PZT 95/5铁电陶瓷

**关键发现:**
1. **失效分布:**
   - 统计了万片铁电陶瓷失效概率与脉冲次数的分布关系
   - 击穿概率与脉冲加压次数曲线呈现典型的**浴盆曲线分布**

2. **可靠性分析:**
   - 经历10次脉冲高压测试合格的样品，第11次失效概率极低
   - 开展了万次以上脉冲耐压后压电常数与电滞回线测试

3. **失效机制:**
   - 高电场下的击穿问题
   - 脉冲电场下的疲劳失效

**C88应用价值:**
- PZT陶瓷的失效模式分析
- 浴盆曲线可用于寿命预测
- 脉冲耐压测试方法可参考

---

## 7. CNKI论文 — 压电陶瓷超声雾化换能器电极保护

### 7.1 论文信息

**标题:** 压电陶瓷超声雾化换能器的电极保护  
**期刊:** 工业技术与职业教育 (CNKI)  
**URL:** https://hbtc.cbpt.cnki.net/portal/journal/portal/client/paper/efe3865f5d005ce52174b6f2595d9cac

### 7.2 研究内容

**关键成果:**
- 对超声雾化换能器银电极的保护进行了讨论
- **开发出使用寿命长达1万小时的雾化换能器**
- 关键词：超声波、雾化换能器、使用寿命

**C88应用价值:**
- 直接相关：1万小时寿命目标
- 电极保护技术（银电极防腐蚀）
- 可联系作者获取详细技术方案

---

## 8. 中国专利

### 8.1 CN107185766A — 压电陶瓷超声雾化片

**专利号:** CN107185766A  
**标题:** 一种压电陶瓷超声雾化片  
**URL:** https://patents.google.com/patent/CN107185766A/zh

**技术特点:**
- 包括第一金属片、第二金属片和压电陶瓷片
- 压电陶瓷片的中部设有通孔
- 第一金属片位于压电陶瓷片的下方
- 第二金属片位于压电陶瓷片的上方

**C88应用价值:**
- 雾化片结构设计参考
- 可规避专利或参考设计

### 8.2 CN213762577U — 双面阶梯式孔径分布雾化片

**专利号:** CN213762577U  
**标题:** 双面阶梯式孔径分布的雾化片  
**URL:** https://patents.google.com/patent/CN213762577U/zh

**技术特点:**
- 金属膜片包含第一金属膜片和第二金属膜片
- 环形压电陶瓷片夹设于两片金属膜片之间
- 构成储液腔
- 双面阶梯式孔径分布

**C88应用价值:**
- 微孔雾化片设计参考
- 阶梯式孔径可优化雾化效率

### 8.3 CN215819888U — 电子烟用超声雾化片

**专利号:** CN215819888U  
**标题:** 电子烟用超声雾化片及电子烟  
**URL:** https://patents.google.com/patent/CN215819888U/zh

**技术特点:**
- 压电陶瓷片面积不超过250mm²
- 厚度不超过0.8mm
- 固有频率不低于1.5MHz
- 额定功率或驱动功率不超过8W

**C88应用价值:**
- 小型化雾化片设计
- 低功耗设计参考

### 8.4 CN117884297B — 复合型压电陶瓷微孔雾化片

**专利号:** CN117884297B  
**标题:** 一种复合型压电陶瓷微孔雾化片  
**URL:** https://patents.google.com/patent/CN117884297B/zh

**技术特点:**
- 包括内压电陶瓷环、外压电陶瓷环和金属薄片
- 进液口尺寸：30~50µm
- 出液口尺寸：1~5µm

**C88应用价值:**
- 复合型设计可提升雾化效率
- 微孔尺寸设计参考

---

## 9. 中国压电陶瓷研究总结

### 9.1 主要研究方向

1. **无铅压电陶瓷** (清华、上交、中南)
   - KNN基无铅压电陶瓷
   - 性能已媲美或超过PZT
   - 环保要求推动

2. **电极保护技术** (CNKI论文)
   - 银电极防腐
   - 1万小时寿命实现

3. **雾化片结构设计** (中国专利)
   - 双面阶梯式孔径
   - 复合型压电陶瓷
   - 小型化、低功耗

4. **失效分析与可靠性** (物理学报)
   - 浴盆曲线失效分布
   - 脉冲耐压测试
   - 寿命预测模型

### 9.2 关键发现

1. **无铅压电陶瓷性能已超PZT:**
   - 上海交大郭益平课题组：d33提升20%，Qm提升80%
   - 雾化片性能已超过同类PZT雾化片

2. **1万小时寿命已实现:**
   - CNKI论文：电极保护技术
   - 关键是银电极防腐

3. **缺陷工程可提升性能:**
   - 清华：氧空位设计，Qm提升60%
   - 上交：锰掺杂，综合性能优化

4. **浴盆曲线适用于压电陶瓷:**
   - 物理学报：失效分布符合浴盆曲线
   - 可用于寿命预测

### 9.3 C88应用建议

1. **材料选择:**
   - 考虑无铅压电陶瓷（环保趋势）
   - 或高性能PZT（成熟技术）

2. **电极保护:**
   - 参考CNKI论文的电极保护技术
   - 或采用镍电极（90年代发明）

3. **寿命预测:**
   - 使用浴盆曲线模型
   - 加速老化测试（阿伦尼乌斯模型）

4. **合作机会:**
   - 联系清华大学王轲研究员
   - 联系上海交大郭益平教授
   - 联系中南大学周学凡副研究员

---

## 10. 参考文献汇总

### 10.1 学术论文

1. Wang et al. "Deciphering the phase transition-induced ultrahigh piezoresponse in (K,Na)NbO3-based piezoceramics" Nature Communications (2022). DOI: 10.1038/s41467-022-31158-x

2. Liu et al. "Isolated-Oxygen-Vacancy Hardening in Lead-Free Piezoelectrics" Advanced Materials (2022). DOI: 10.1002/adma.202202558

3. Wang et al. "Defect Dipole-Induced Fatigue Strain Recoverability in Lead-Free Piezoelectric Ceramics" Advanced Functional Materials (2023). DOI: 10.1002/adfm.202306416

4. Guan et al. "Superior Electromechanical Compatibility in Lead-free Piezoceramics with Mobile Transition-Metal Defects" ACS Applied Materials & Interfaces (2023). DOI: 10.1021/acsami.3c12068

5. Yin et al. "Modulation of piezoelectricity and mechanical strength in piezoelectric composites based on N0.5B0.51T-BNT nanocubes towards human-machine interfaces" Nano Energy (2023). DOI: 10.1016/j.nanoen.2023.109044

6. 铁电陶瓷脉冲耐压失效分布及耐压可靠性. 物理学报 (2024). DOI: 10.7498/aps.73.20231354

7. 压电陶瓷超声雾化换能器的电极保护. 工业技术与职业教育 (CNKI).

8. 压电雾化装置的研究进展. 南京航空航天大学学报 (2020).

9. 韩光超. 压电式超声雾化降尘系统设计及其试验研究. 机械科学与技术 (2024), 43(2): 258-264.

### 10.2 中国专利

10. CN107185766A — 一种压电陶瓷超声雾化片  
11. CN213762577U — 双面阶梯式孔径分布的雾化片  
12. CN215819888U — 电子烟用超声雾化片  
13. CN117884297B — 复合型压电陶瓷微孔雾化片

### 10.3 研究团队联系方式

**清华大学 王轲研究员:**
- 单位：清华大学材料学院
- 研究方向：无铅压电陶瓷
- 论文：Nature Communications, Advanced Materials

**上海交大 郭益平教授:**
- 单位：上海交通大学材料科学与工程学院
- 金属基复合材料国家重点实验室
- 论文：Advanced Functional Materials, ACS AMI, Nano Energy

**中南大学 周学凡副研究员:**
- 单位：中南大学粉末冶金研究院
- 团队主页：http://www.csupiezogroup.com/
- 电话：18073153541
- 研究方向：无铅压电陶瓷材料与器件，超声雾化片

**中国地质大学 韩光超教授:**
- 单位：中国地质大学（武汉）机械学院
- 研究方向：压电式超声雾化降尘系统

---

**End of Chinese Academic Research Papers**

**知识库更新完成:**
- 01-压电物理 (14KB)
- 02-防腐涂层 (19KB)
- 03-驱动电路 (15KB)
- 04-寿命测试 (14KB)
- 05-失效模式 (20KB)
- 06-中国技术补充 (13KB)
- **07-中国学术论文 (新增, ~15KB)**

**总计：7章，~110KB技术知识**

**周六SWT拜访可以引用:**
- 清华大学Nature Communications论文
- 上海交大Advanced Functional Materials论文
- CNKI 1万小时寿命论文
- 中国专利设计参考


