UV-Kammer: Dielektrische Korrosion und Feuchtigkeitseintritt von Hochspannungskondensatoren

UV-Kammer: Dielektrische Korrosion und Feuchtigkeitseintritt von Hochspannungskondensatoren

Abstract

A 10 kV capacitor in a curing chamber sees 200–450 nm radiation, ozone, 85% RH and 60 °C at the same time. The sintered ceramic is inert under that flux; the epoxy seal, the silver termination and the creepage path are not. This article gives the photon energy and surface conductivity arithmetic, four field failure modes, a 1000 h qualification matrix, and the materials that hold a 4700 pF part inside ±5%.

1. Einleitung

A UV curing machine places a high-voltage capacitor in warm, humid, oxidizing air that is also full of ultraviolet light. Direct and scattered radiation from 200 to 450 nm reaches the trigger, ballast and filter stages, and the lamp housing adds ozone and heat. Body temperature near 60 °C holds for a full shift.

The ceramic core of a Class I part is a sintered oxide, and it survives that exposure. The parts that fail are the interfaces around it: the epoxy that seals the body, the metal that connects the electrodes, and the creepage path between terminals. Each interface is attacked by a different agent.

Returns from humid UV lines show four signatures. A 10 kV trigger capacitor grows a black carbon track in 12 to 18 months, a silver dendrite bridges a termination, a filter drops 8 dB of rejection after water uptake, and a crack opens at the lead exit. A dry bench test shows none of them.

This article gives the mechanism arithmetic, the four field failure modes, the qualification matrix, and the material choices that hold a 4700 pF, 10 kV part inside ±5% for the life of the lamp.

2. Kernmechanismus

2.1 Photon energy and chain scission

A UV photon carries an energy that depends on wavelength alone. Above the carbon-carbon bond energy, the radiation breaks polymer chains instead of heating them.

E_ph = h · c / λ

where E_ph is the photon energy, h is the Planck constant, c the speed of light and λ the wavelength. At 300 nm, E_ph works out at 4.1 eV, which exceeds the 3.6 eV carbon-carbon bond energy of the epoxy that seals the capacitor body.

2.2 Mass loss rate in the encapsulant

The rate at which the encapsulant loses mass follows an Arrhenius form scaled by the UV flux.

dm/dt = k · Φ_UV · exp(−E_a / (R · T))

where dm/dt is the polymer mass loss rate, k is a material constant, Φ_UV is the UV flux, E_a is the activation energy, R is the gas constant and T is the absolute temperature. Doubling the flux roughly doubles the loss rate, so a two-tube lamp ages the seal about twice as fast as a single-tube lamp.

2.3 Surface conductivity and ionic migration

A degraded surface turns hydrophilic, and absorbed water dissolves the ionic contamination left by flux residue and by atmospheric sulfate. Surface conductivity rises with relative humidity.

σ_s = σ_0 · exp(α · RH)

where σ_s is the surface conductivity, σ_0 its dry value, α the humidity coefficient and RH the relative humidity. At 85% RH the surface conductance reaches about 1000× its dry value, which turns a clean 20 kV creepage path into a leakage sink. Silver then travels at 0.1–1 mm per 1000 h at 10 kV/cm in humid air and grows a dendrite from anode to cathode.

2.4 Parameter drift on the same surface

Water uptake shifts both the capacitance and the loss of the part. The two relations below tie that drift to the pulse energy and to the filter tuning.

E = 0.5 · C · V²

where E is the stored pulse energy, C is the capacitance and V is the charging voltage. With C in nF and V in kV the result comes out in mJ, so a 4.7 nF part at 10 kV stores 235 mJ per pulse.

f_r = 1 / (2 π √(L_s · C))

where f_r is the self-resonant frequency, L_s is the equivalent series inductance and C is the capacitance. At L_s = 8 nH and C = 4700 pF the resonance sits at 26 MHz. Water taken up at 0.1% by weight raises the effective permittivity, shifts C by several percent and moves that resonance; a filter trimmed for 30 dB of rejection falls to 22 dB.

Fig. 1 — UV photons break the epoxy seal, humidity raises surface leakage a thousandfold, and silver migration carbonizes the 10 kV creepage path.

3. Schwachstellen im Engineering

Pain Point 1: Tracking flashover on an epoxy-sealed body

A 10 kV trigger capacitor in a UV chamber grows a carbon track between its terminals in 12 to 18 months. UV-degraded epoxy plus humidity plus voltage carbonizes the leakage film week by week. The lamp starts to miss ignitions and then flashes over across the body. A glass-glazed construction leaves no polymer surface to carbonize.

Pain Point 2: Silver migration at the terminations

Silver-palladium terminations in coastal or sulfidizing air corrode, and the released silver ions then travel under the DC field. Termination corrosion is the most common field return for HV capacitors in humid UV lines, and it stays invisible until the part fails short. A 1 pF leak becomes a short circuit in one night.

Pain Point 3: Moisture-induced parameter drift

A capacitor that takes up 0.1% water by weight shifts its capacitance by several percent and raises tanδ. A filter trimmed for 30 dB of attenuation drops to 22 dB, and the machine shows optical ripple that the dry laboratory cannot reproduce.

Pain Point 4: Delamination at the lead exit

Differential thermal expansion and UV embrittlement crack the seal around the lead. The crack admits moisture, and partial discharge starts in the void. From that point the capacitor follows a slow path to failure no matter how well the ceramic itself performs.

4. Entwurfsschritte

Strategy 1: Choose an inorganic dielectric for UV-exposed sites

Verlegen Sie N4700 (T3M) strontium titanate and a body whose exposed surfaces are glass-sealed or ceramic-glazed instead of epoxy-domed. A Folienkondensator with an organic dielectric is the weaker choice at this site, because the photons that crack the epoxy also attack the polymer film.

Strategy 2: Qualify with a humidity and UV matrix

Run 1000 h at 85 °C / 85% RH with bias applied, then 500 h of 200–400 nm exposure at 10× the operating flux. Measure capacitance, tanδ and leakage current every 100 h. Accept a part only when tanδ stays at or below 0.1% and the leakage current shows no upward trend.

Strategy 3: Specify corrosion-resistant terminations

Use nickel-barrier or protected screw-terminal terminations, and ask for sulfur-resistance and salt-spray data with the quotation. At the harshest sites the sealed screw-terminal construction holds the electrode stack away from the chamber atmosphere.

Strategy 4: Control the local atmosphere

Seal the high-voltage compartment, fit a desiccant breather, and keep the capacitor body above the dew point of the cabinet. A 10 °C rise in body temperature halves the equilibrium surface moisture, so a part that runs warm and dry ages slower than a part that runs cool and wet.

Strategy 5: Add creepage margin and a UV-stable coating

Increase the terminal-to-terminal creepage distance by 30–50% over the standard spacing for the voltage class. Then coat the capacitor and the adjacent board with a UV-stable silicone conformal coating that does not yellow or crack under the lamp.

Strategy 6: Screen partial discharge after humidity preconditioning

Moisture lowers the PD inception voltage, so a part that passes a dry test can fail in the first rainy month. Run the PD test after humidity preconditioning and accept only parts below 5 pC at 1.2× the rated voltage.

Komponentenparameter	Typischer Wert	Testbedingung
Nennspannung (HVCT8G-Serie)	10 kV DC	Hi-Pot 1.5× für 60 s
Kapazität	4700 pF ±20% (472M)	1 MHz, 1 V<sub>eff</sub>
ESR bei 1 MHz	unter 50 mΩ	25 ° C
Schleifeninduktivität (E40-Sehne)	unter 8 nH	40 mm Anschlussabstand
Isolationswiderstand	über 10 GΩ bei 500 V Gleichstrom	25 °C, 60 % relative Luftfeuchtigkeit
Temperaturbereich	−40 °C bis +125 °C	IEC 60068-2-Reihe
Impuls-dv/dt-Fähigkeit	über 50 kV/µs	bei Nennspannung

5. HVC Class I capacitors for UV chamber service

Three requirements drive the choice at this node: a dielectric that ignores the UV flux, a seal that keeps water out of the interface, and a termination that resists sulfidization. HVC builds N4700 parts for that set.

Sechs Parameter, die hier von Bedeutung sind

• N4700 (T3M) strontium titanate holds tanδ at or below 0.1% at 1 kHz, and the sintered surface is photochemically inert under 200–450 nm flux.
• ±30 ppm/°C holds the 4700 pF value and the 26 MHz self-resonance steady from −40 °C to +125 °C.
• Ratings from 1 kV to 100 kV cover the trigger, ballast and filter stages, with sealed and glass-glazed versions at every class.
• PD below 5 pC at 1.2× rated voltage on every part, including after humidity preconditioning, closes the moisture-assisted PD path.
• Nickel-barrier and protected HVCT8G screw-terminal terminations resist silver migration and sulfidization, verified by salt-spray and mixed-gas testing.
• Field validation: an HVC 40 kV / 441 screw-type capacitor replaced a first-tier brand part in high-voltage field-service programs, and the same glass-sealed Class I construction matches the electrical parameters of Vishay Radialanschluss-Baureihen von HVCC und KEMET.
• Above 1 kV the site is available as a series stack, and HVC supplies matching Bemessung von Widerständen Der Stapel behält also seine Spannungsteilung auch nach thermischen Zyklen bei.

Lieferung und Support

Standard parts ship in 4–8 weeks and matched stacks in 6–10 weeks. HVC applications engineering reviews the UV flux, the humidity profile and the ambient gases for the site.

Send the flux at the capacitor body, the cabinet humidity, the gases present and the creepage drawing. HVC returns a dielectric and termination recommendation with data from the humidity and UV qualification matrix, and glass-sealed samples in two weeks.

6. Häufig gestellte Fragen

Q1: Why does a capacitor pass a dry PD test and fail in the chamber?

Because moisture lowers the PD inception voltage. Water absorbed into the surface film and into a void at the lead exit lets the first discharge start well below the dry inception level. A part screened dry at 5 pC can exceed 50 pC after 1000 h at 85% RH, which is why the PD screen belongs after preconditioning.

Q2: How fast does silver migration grow at 10 kV/cm?

Field and laboratory data put silver growth at 0.1–1 mm per 1000 h in humid air at 10 kV/cm. A 4 mm termination gap therefore bridges in 4,000 to 40,000 h, and the dendrite resistance collapses the part within seconds of the bridge closing.

Q3: Which dielectric survives a 200–450 nm flux?

A sintered Class I oxide. N4700 (T3M) strontium titanate has a bond structure that UV photons at 4.1 eV cannot break, so the ceramic body shows no mass loss. Polymer encapsulants and organic film dielectrics absorb the same photons and degrade.

Q4: How much capacitance drift does 0.1% water uptake cause?

Several percent, plus a rise in tanδ. The larger effect is on the filter: a 30 dB stopband falls to 22 dB, and the self-resonance of a 4700 pF part with 8 nH of series inductance moves off 26 MHz.

Q5: What creepage distance is needed at 10 kV in a UV chamber?

Take the standard spacing for the 10 kV class and add 30–50%. Then coat the surface with a UV-stable silicone, because the creepage distance alone does not help once the surface film becomes conductive at 85% RH.

Q6: How is a UV-resistant capacitor qualified?

Run 1000 h at 85 °C / 85% RH with bias, then 500 h at 200–400 nm and 10× the operating flux. Measure capacitance, tanδ and leakage every 100 h, and finish with a PD test below 5 pC at 1.2× rated voltage.

7. Referenzen

1. E. Kuffel, W. S. Zaengl, J. Kuffel, Grundlagen der Hochspannungstechnik, 2. Aufl., Newnes, 2000.
2. N. Mohan, TM Undeland, WP Robbins, Leistungselektronik: Wandler, Anwendungen und Design, 3. Auflage, Wiley, 2003.
3. MIL-HDBK-217F Mitteilung 2, Zuverlässigkeitsvorhersage elektronischer Geräte 1995.
4. IEC 60384-1:2018, Festkondensatoren für den Einsatz in elektronischen Geräten – Teil 1: Allgemeine Spezifikation.
5. DN Donahoe, M. Pecht, IK Lloyd und S. Ganesan, Feuchtigkeitsbedingte Degradation von Mehrschicht-Keramikkondensatoren, Microelectronics Reliability, Bd. 46, Nr. 2-4, S. 400-408, 2006. https://doi.org/10.1016/j.microrel.2005.05.008.
6. IEC 60270:2025, Hochspannungsprüfverfahren – Ladungsbasierte Messung von Teilentladungen, 4. Aufl. Genf: IEC, 2025.
7. Entwicklung von BME X7R Mehrschicht-Keramikkondensatoren mit hoher Kapazität und hoher Spannung, KEMET Electronics (CARTS USA 2011 Proceedings), https://www.mouser.jp/pdfdocs/2011-03CARTS-DevelopmentofHighCapacitance-HighVoltageBMEX7RMLCCapacitors.pdf .
8. AJ Moulson und JM Herbert, Elektrokeramik: Materialien, Eigenschaften, Anwendungen, 2. Aufl. Chichester: Wiley, 2003.
9. GB/T 7354-2018, 高电压试验技术 局部放电测量. Titel: 2018.
10. Verwendung von Keramikkondensatoren 3 kV (Stabilitätsdiskussion C0G/NP0 vs X7R/Y5S), Forum Retrotechnique, https://forum.retrotechnique.org/t/utilisation-des-condensateurs-ceramiques-3kv/115858 .
11. HVC-Anwendungsentwicklungsteam Hochspannungs-Keramikkondensator Technische Ressourcen, https://www.hv-caps.com