Conception d'allumage étanche à l'humidité pour la découpe plasma en ateliers humides

Conception d'allumage étanche à l'humidité pour la découpe plasma en ateliers humides

Abstract

Machines that strike perfectly in a dry lab misfire at dawn in a humid shop, and the fault follows the weather instead of the parts. This article quantifies surface leakage, the capacitance and loss shift from absorbed moisture, the detuning of the ignition tank and the humidity shift of gas breakdown, then gives four pain points, six design steps and the sealed capacitor data that takes moisture off the critical path.

1. Introduction

A plasma cutter in a coastal shop, a humid foundry or an unheated warehouse meets water vapour on every surface. Moisture films on high-voltage surfaces shorten the effective breakdown distance and start surface tracking, condensation on the tank and transformer lowers insulation resistance, absorbed moisture in windings raises dielectric loss, and humid air in the gap changes the breakdown statistics.

The pain shows up as intermittency. Humidity-driven misfires follow weather and season instead of component wear, so the engineering lab cannot reproduce them. Field engineers rotate gas bottles, electrodes and whole ignition modules, and the fault returns with the next wet week.

Automated cells carry the highest cost. A dawn misfire interrupts a nesting program and scraps the work in process. In high-duty shops the retries, the consumables spent on failed starts and the lost throughput add up across every humid season.

This article puts the moisture path into numbers: the leakage across a wet creepage path, the capacitance and loss shift inside the dielectric, the detuning of the ignition tank, and the breakdown voltage the gap needs at 95% relative humidity.

2. Mécanisme de base

2.1 Leakage across a wet surface

A water film dissolves ionic contamination on an insulating surface and turns it into a conductive layer.

i_leak = (σ_surf × V × w) / d

where i_leak is the leakage current, σ_surf is the surface conductance, V is the applied voltage, w is the wet-path width and d is the creepage distance.

Surface conductance climbs orders of magnitude above 60% RH. On a 25 mm creepage path 10 mm wide at 15 kV, a clean surface at 1 × 10⁻¹⁴ S carries 60 nA, and the same geometry with ionic contamination at 1 × 10⁻⁺ S carries 60 µA. The second number is the one that matters.

2.2 Dry-band arcing

Leakage current heats the film, and the power it leaves behind follows the surface resistance.

R_surf = d / (σ_surf × w)

where R_surf is the wet-path resistance, d is the creepage distance, σ_surf is the surface conductance and w is the path width.

P_leak = i_leak² × R_surf

where P_leak is the power dissipated in the film, i_leak is the leakage current and R_surf is the film resistance. The dry path measures 2.5 × 10¹⁴ Ω and burns 0.9 mW at 95% RH. Contamination takes the same path to 2.5 × 10⁸ Ω, and the dissipation reaches 0.9 W, which dries a narrow band, raises the local voltage gradient and starts the arcing that carbonises the surface.

2.3 Capacitance, loss and resonance

Absorbed moisture raises the permittivity of the dielectric and the loss factor with it, which moves the resonance of the ignition tank.

Δf_res / f_res = −½ × (ΔC / C)

where Δf_res is the shift in tank resonance, f_res is the tuned resonance, and ΔC / C is the fractional capacitance change.

A film dielectric that fills 0.05% of its volume with water raises its capacitance by 1.8% and its loss factor from 0.02% to 0.15%. The tank resonance then falls 0.9%, which on a 1.896 MHz tank is 17 kHz. The loss rise is smaller than the detuning in its consequences, but it also adds local heating on a surface that is already wet.

2.4 What detuning costs at the gap

An ignition tank driven away from resonance delivers less voltage, and the penalty scales with Q.

V_tank / V_tuned = 1 / √(1 + (2Q × Δf / f)²)

where V_tank is the delivered tank voltage, V_tuned is the voltage at exact resonance, Q is the loop quality factor and Δf / f is the fractional frequency offset.

A loop at Q 30 with a 0.9% offset returns 0.88. A 10 kV tank therefore delivers 8.8 kV to the gap instead of 10 kV, and the sealed part that holds capacitance inside 0.2% returns 0.998, which is 9.98 kV.

2.5 What the humid gap needs

Water vapour changes the avalanche coefficients, so humid air needs more voltage for the same spacing.

V_brk(RH) ≈ V_brk(0) × (1 + β × RH)

where V_brk(RH) is the breakdown voltage at relative humidity RH, V_brk(0) is the dry-air value and β is a positive coefficient near 2 × 10⁻³ per %RH for air gaps.

A gap that breaks down at 8.20 kV at 20% RH needs 9.38 kV at 95% RH, which is 14% more. That figure corrects the common description of a shift of a few percent: with β at 2 × 10⁻³ per %RH the 75-point span from a dry season to a wet one costs 14%.

2.6 The margin that closes

The stored energy per strike follows the usual pair of terms.

E = ½ × C × V²

where E is the stored pulse energy, C is the tank capacitance and V is the charge voltage. With C in nanofarads and V in kilovolts the result is in millijoules, so 4.7 nF at 10 kV gives 235 mJ and not 235 J.

Dry season: a 10 kV tank delivers 9.98 kV against a gap demand of 8.20 kV, which is 22% of margin. Wet season with an unsealed part: 8.8 kV against 9.38 kV, which is 6% short and a misfire. Sealing the dielectric removes the detuning, and the 8% pre-charge boost that the firmware adds above 70% RH takes the tank to 10.78 kV, which is 15% of margin.

Fig. 1 — A dry tank clears the gap demand by 22%; at 95% RH the detuned tank falls 6% short of what the wet gap needs.

3. Points faibles en matière d'ingénierie

Pain Point 1: Surface tracking on high-voltage insulators

Creepage sized for clean dry air is too short in a humid shop. Tracking develops over months on the transformer bobbin, the coupling capacitor and the torch head, and once the carbonised path is in place the ignition voltage discharges through it instead of across the gas gap. The path does not recover when the air dries.

Pain Point 2: Moisture absorption inside the tank capacitor

A condensateur à film and a porous ceramic body both take up water. Capacitance rises 1.8%, the loss factor rises from 0.02% to 0.15%, the tank is 17 kHz off tune, and the delivered voltage falls 12%. Visual inspection finds nothing, because the part looks new.

Pain Point 3: Condensation-driven partial discharge

Water on a capacitor termination and on winding surfaces lowers the local discharge inception voltage. Sustained discharge then erodes the insulation and radiates broadband noise that reaches the control electronics at the moment the machine is already failing to start, which hides the real cause.

Pain Point 4: The dew-point trap in a sealed enclosure

A sealed enclosure keeps water out and keeps humidity in. When the machine cools overnight, internal surfaces fall below the dew point and condensation wets every part inside. Heaters and fans added afterwards usually break the EMC gasket in the process.

4. Étapes de conception

Strategy 1: Remove the moisture-absorbing element

Spécifiez un N4700 Class I capacitor with a dense vitrified dielectric and sealed terminations, so capacitance and loss stay put from 20% to 95% RH. The 0.2% capacitance window keeps the tank resonance inside 0.1% and the delivered voltage above 9.9 kV.

Strategy 2: Size creepage and clearance for the wet case

Use the IEC 60664-1 pollution degree 3 creepage table for the ignition voltage and add 20% beyond the table value. Add anti-tracking geometry to the transformer bobbin and the capacitor holders: ribs that lengthen the surface path and slots that break it, so a wet film cannot bridge the gap in a straight line.

Strategy 3: Control the enclosure microclimate

Fit a 5 to 15 W heater that holds internal surfaces 5 to 10 °C above the dew point during standby, and vent the enclosure through a breathable membrane instead of sealing it hermetically. That combination keeps the interior dry without trapping the water that condenses at dawn.

Strategy 4: Pot or coat the high-voltage assembly

Encapsulate the transformer secondary and the capacitor terminations in a void-free, low-absorption resin. Check afterwards that the potting has not moved the winding resonances by more than the design tolerance, because a resin with a high permittivity changes the layer capacitance that sets the 73.6 kHz mode.

Strategy 5: Compensate in firmware against measured humidity

Read a humidity sensor or a dew-point estimate and raise the pre-charge voltage by up to 8% when RH passes 70%. That covers part of the 14% rise in gap demand, and the sealed dielectric removes the 12% detuning loss, so the two together restore the margin instead of spending consumable life on every strike.

Strategy 6: Qualify with a condensation cycle test

Cycle the complete ignition module 100 times between 25 °C at 20% RH and 35 °C at 95% RH with a condensation phase, and fire the igniter at the end of each cold phase. Require 100% first-attempt ignition across the whole cycle, and repeat the test after any change to potting, sealing or creepage geometry.

Paramètre de composant	Valeur typique	Condition de test
Tension nominale (HVCT8G séries)	10 kV DC	haute tension 1.5× pendant 60 s
Capacitance	4700 pF ±20% (472M)	1 MHz, 1 V efficace
Capacitance against humidity	inside 0.2%	20% to 95% RH, 25 °C
ESR à 1 MHz	inférieur à 50 mΩ	-
La resistance d'isolement	supérieure à 10 GΩ à 500 V CC	25 °C, 60 % HR
Décharge partielle	en dessous de 5 pC	at rated voltage, 95% RH
Plage de température	−40 °C à +125 °C	Série IEC 60068-2

5. HVC N4700 (T3M) capacitors for humid shops

Reliability in a wet season needs parts that ignore the water. HVC builds the N4700 (T3M) Class I family for that duty, with the résistance à couche épaisse et la résistance d'échantillonnage for the leakage monitor, from a 6,000 m² owned factory in Dongguan across a 1 kV to 100 kV range.

Six paramètres importants ici

• Negligible moisture absorption. The vitrified dielectric and sealed terminations hold capacitance and loss factor from dry air to saturated air, so the ignition tuning measured in the lab survives a wet week on the shop floor.
• Loss that does not climb with the weather. A dissipation factor below 0.1% keeps local heating out of the body, which removes the heat source that starts dry-band arcing on a wet component surface.
• Discharge below 5 pC at high humidity. A dense dielectric suppresses internal and surface discharge, so the capacitor does not become the tracking or EMI source inside a humid enclosure.
• Voltage margin without oversizing. Ratings from 1 kV to 100 kV let the design keep the generous margins that pollution degree 3 creepage needs while the package still fits the ignition module.
• Stable temperature behaviour. A ±30 ppm/°C coefficient keeps the ignition resonance in place between a warm afternoon and a cold wet night, which decouples the misfire rate from the weather.
• Proven in humidity-controlled plants. The 40 kV 441-series doorknob capacitors replaced first-tier brand parts in high-voltage field-service programs, and a références croisées against those parts shows the humidity drift a design would have to absorb.

Livraison et support

Standard parts ship in 4 to 8 weeks and matched ignition sets in 6 to 10 weeks. HVC applications engineering supports creepage layout, enclosure microclimate design, potting qualification and condensation-cycle test setup. Send the ignition voltage, the RH range and the enclosure description, and HVC returns a capacitor selection sheet with humidity-stability data and PD reports within five working days.

6. Questions fréquentes

Q1: How much leakage does a wet creepage path carry?

On a 25 mm path 10 mm wide at 15 kV, a clean surface at 1 × 10⁻¹⁴ S carries 60 nA. The same geometry with ionic contamination at 1 × 10⁻⁺ S carries 60 µA, and the film then dissipates 0.9 W, which is enough to dry a band and start the arcing that carbonises the surface.

Q2: Why do humid misfires look like component faults?

Because moisture acts on three independent mechanisms at once: surface leakage, dielectric absorption and gap breakdown. Each one produces the same symptom of a failed first attempt, so replacing the gas, the electrode or the ignition module cures nothing and the fault returns with the next wet week.

Q3: How much does moisture shift the tank tuning?

A film dielectric that takes up 0.05% water by volume gains 1.8% capacitance, and the resonance falls by half that, 0.9%. On a 1.896 MHz tank that is 17 kHz. A loop at Q 30 then delivers 0.88 of its tuned voltage, so 10 kV becomes 8.8 kV.

Q4: How much more voltage does the gap need at 95% RH?

With a humidity coefficient near 2 × 10⁻³ per %RH, a gap that breaks down at 8.20 kV at 20% RH needs 9.38 kV at 95% RH. That is 14% more, not the few percent that the phrase breakdown shift often suggests.

Q5: What size heater and vent does the enclosure need?

A 5 to 15 W heater that holds internal surfaces 5 to 10 °C above the dew point during standby, plus a breathable membrane vent. Sealing hermetically without a vent traps the humidity that condenses overnight, and the condensation then wets every surface inside the enclosure.

Q6: Which capacitor holds its value through a wet season?

An N4700 Class I part with a vitrified dielectric, sealed terminations, a ±30 ppm/°C coefficient and discharge below 5 pC at rated voltage. It holds capacitance inside 0.2% from 20% to 95% RH, which keeps the delivered tank voltage above 9.9 kV and leaves the firmware boost to cover the gap demand.

7. Références

1. E. Kuffel, WS Zaengl, J. Kuffel, Principes fondamentaux de l'ingénierie haute tension, 2e éd., Newnes, 2000.
2. Mohan, N., Undeland, TM et Robbins, WP, Électronique de puissance : convertisseurs, applications et conception, 3e éd. Hoboken, NJ : John Wiley & Sons, 2003.
3. Avis MIL-HDBK-217F 2, Prédiction de la fiabilité des équipements électroniques 1995.
4. CEI 60664-1:2020, Coordination de l'isolation des équipements des réseaux d'alimentation basse tension - Partie 1 : Principes, exigences et essais, 3e éd. Genève : CEI, 2020.
5. CEI 60384-1:2021, Condensateurs fixes pour équipements électroniques - Partie 1 : Spécifications générales, 6e éd. Genève : CEI, 2021.
6. DN Donahoe, M. Pecht, IK Lloyd et S. Ganesan, Dégradation des condensateurs céramiques multicouches induite par l'humidité, Fiabilité en microélectronique, vol. 46, n° 2-4, pp. 400-408, 2006. https://doi.org/10.1016/j.microrel.2005.05.008.
7. M. Xiao et al., Recherche sur les caractéristiques de décharge partielle et d'auto-réparation des films métallisés de condensateurs, dans les actes de la 5e Conférence internationale de l'IEEE sur les diélectriques (ICD), p. 1-4, 2024. https://doi.org/10.1109/icd59037.2024.10613348
8. Miniaturisation des condensateurs céramiques multicouches haute tension BME X7R pour applications automobiles, J. Bultitude et al., KEMET Electronics, https://www.mouser.lu/pdfdocs/KEMET-TechnicalPaperBultitude_AECAutoRel_HVMiniature150410_Final-2.pdf .
9. États-Unis 6,760,215, Condensateur à seuil de claquage à haute tension, DF DeVoe (lié au Presidio), accordé en 2004, https://www.freepatentsonline.com/6760215.html .
10. GB/T 6346.1-2024, 电子设备用固定电容器 第1部分：总规范 (Condensateurs fixes destinés à être utilisés dans les équipements électroniques – Partie 1 : Spécification générique). Pékin : 国家市场监督管理总局, 2024.
11. Équipe d'ingénierie des applications HVC, Condensateur céramique haute tension Ressources techniques, https://www.hv-caps.com