等离子切割机绝缘保护的点火尖峰抑制

等离子切割机绝缘保护的点火尖峰抑制

抽象

Arc ignition is the most violent electrical event in a plasma cutting machine, and the subsystem built to make high voltage is the one that destroys the parts around it. This article gives the trapped-energy equation, the spike amplitude bound, the snubber damping target, the cable reflection rule, four field pain points, six protection steps and the worst-case test that closes the design.

1. 引言

Ignition is the most violent event in a plasma cutting machine. A capacitor charged to several kilovolts dumps its energy into an inductance-limited loop in microseconds.

The voltage excursions that follow reach two to three times the steady tank voltage, the current ramps faster than 1 kA/µs, and the radiated field couples into every wire within a metre of the torch.

Those spikes cause most of the switch failures, transformer winding breakdowns and control-board faults in plasma supplies. The ignition subsystem exists to make high voltage, and it damages the electronics around it while doing so.

This article gives the three energy sources behind the spike, the equations that bound amplitude and energy, four pain points, six suppression steps and the worst-case test that proves the design.

2. 核心机制

2.1 Trapped field energy

When the gap breaks down, the energy left in the tank and the energy in the transformer's magnetic field both need a path. The trapped field energy is the larger term.

E_trapped = ½ L_σ i_peak²

where E_trapped is the trapped energy in joules, L_σ is the leakage inductance referred to the primary, and i_peak is the primary current at breakdown. A 10 µH leakage inductance carrying 217 A holds 235 mJ, exactly the energy stored in a 4.7 nF tank at 10 kV.

2.2 Spike amplitude at the switch node

When the arc extinguishes or the switch opens, the trapped energy charges the parasitic capacitance of the switch node, and the resulting voltage follows an energy balance.

V_spike = i_peak √(L_σ / C_par)

where V_spike is the peak node voltage, and C_par is the parasitic capacitance at the switch node including the clamp capacitor. The same 217 A into 520 pF produces 30 kV, and into 100 pF it produces 68.6 kV. Node capacitance is the design lever, and it is also the number that shrinks when the module gets smaller.

The characteristic impedance of the loop follows from the same two elements.

Z_0 = √(L / C)

where Z_0 is the characteristic impedance, L is the loop inductance and C is the loop capacitance. With L in nanohenries and C in picofarads the expression gives Z_0 in ohms after multiplying by 31.6, so the 10 µH and 4.7 nF pair presents 46 Ω and 217 A at 10 kV.

2.3 The inductive kick during turn-off

The switch's own transition adds a second term while the current falls and the voltage rises.

ΔV = L_loop (di/dt)

where ΔV is the inductive kick, L_loop is the inductance of the commutation loop, and di/dt is the current slope during the fall time. The unit rule is convenient: nanohenries multiplied by kiloamperes per microsecond gives volts. A 50 ns fall interrupting 217 A gives 4.34 kA/µs, and 200 nH of commutation loop turns that into 868 V.

2.4 Cable reflection and snubber damping

The cable between the ignition module and the torch reflects the pulse, and an unmatched load can double or triple the voltage at the torch end.

V_torch ≈ V_cable (1 + Γ), Γ = (Z_load − Z_c) / (Z_load + Z_c)

where V_torch is the voltage at the torch end, V_cable is the incident voltage, Γ is the reflection coefficient, Z_load is the load impedance and Z_c is the cable characteristic impedance. Matching Z_load to Z_c drives Γ to zero and removes the doubling.

The snubber handles the remaining excursion, and its damping ratio sets the overshoot.

overshoot = exp(−π ζ / √(1 − ζ²))

where ζ is the series damping ratio of the snubber loop. A damping ratio of 0.5 gives 16 % overshoot, or 1.16× the step, and 0.8 gives 1.5 %. The target band is 0.5 to 0.8, which keeps the first excursion near 1.2× the steady value.

Fig. 1 — The same 217 A of trapped current gives 30 kV across 520 pF of node capacitance and 68.6 kV across 100 pF.

3. 工程痛点

Pain Point 1: Spike amplitude growing as the module shrinks

Compact layouts use longer, thinner leads and tighter routing, which raises loop inductance while removing the stray capacitance that used to absorb the spike. Both terms move the wrong way at the same time, and the insulation margin falls with them.

Pain Point 2: Avalanche ratings treated as an afterthought

Power switches are chosen on steady-state ratings, and avalanche energy gets whatever margin is left. Ignition duty is a continuous train of avalanche events, so each spike consumes part of a cumulative allowance and the failures appear in the field as unexplained switch deaths.

Pain Point 3: Transformer inter-turn insulation under 20–50 ns edges

The reflected high-frequency content concentrates across the first turns of the primary and the inter-layer dielectric of the secondary. An uncontrolled termination stresses that insulation far beyond its 50 Hz test rating, and the damage accumulates without any visible sign at the terminals.

Pain Point 4: Coupling into the control plane

Fast edges couple through transformer stray capacitance and ground planes into analogue inputs, producing false readings and spurious fault flags. The same edges reach sensor wiring and servo loops, so the machine misbehaves intermittently and the fault cannot be reproduced in the lab.

4. 设计步骤

Strategy 1: Fit a snubber sized from measured parasitics

Measure L_σ and C_par, then size an RC or RCD snubber to absorb the trapped energy. Target a damping ratio between 0.5 and 0.8 so the first voltage excursion stays near 1.2× the steady-state value.

Strategy 2: Terminate the cable to the torch

Add a series resistance equal to the cable characteristic impedance at the torch end, or a parallel clamp. That collapses Γ toward zero and removes the double-travel doubling, which is the largest single voltage term in the chain.

Strategy 3: Clamp the bus close to the switch

适合一个 薄膜电容器 with low ESR across the DC bus, within 10 mm of the switch, so the commutation loop stays small and the di/dt term sees the least possible inductance.

Strategy 4: Choose a snubber capacitor with a low PD floor

指定 N4700 (T3M) Class I parts: a SrTiO3 dielectric with tanδ at or below 0.1 %, a temperature coefficient of ±30 ppm/°C, 1 kV to 100 kV ratings and PD below 5 pC, so the snubber neither adds loss nor becomes a discharge source inside a sealed module.

Strategy 5: Slow the edges at the control boundary

Add series ferrite beads and 10–100 pF shunt capacitors to every signal line that leaves the ignition enclosure, and use a single-point ground star so spike currents cannot flow through the control ground.

Strategy 6: Run the worst-case ignition test

Fire the igniter into a dead short, into an open circuit and into a marginal gap while logging the switch drain voltage, the transformer secondary voltage and the control-plane readings. Freeze the design when all three stay inside their envelopes.

组件参数	典型值	测试条件
额定电压（HVCT8G系列）	10 kV DC	高压1.5倍，持续60秒
电容量	4700 pF ±20% (472M)	1 MHz，1 V rms
1 MHz 下的 ESR	低于 50 mΩ	25°C
回路电感（E40弦）	低于 8 nH	端子间距 40 毫米
绝缘电阻	500V直流电压下电阻大于10GΩ	25℃，60%相对湿度
温度范围	−40 °C 至 +125 °C	IEC 60068-2系列
脉冲dv/dt能力	高于 50 kV/µs	额定电压

5. HVC capacitors for transient suppression

Three properties make a suppression capacitor safe to bury in a sealed module: low loss, a stable clamp threshold and a clean discharge floor. HVC builds N4700 parts for all three.

这里需要考虑的六个参数

• tanδ at or below 0.1 % makes the snubber dissipate its energy in the resistor, so damping stays predictable and the capacitor stays cool.
• ±30 ppm/°C holds the effective clamp voltage of divider and clamp circuits steady, so the protection margin does not shrink on a hot day.
• 1 kV to 100 kV ratings let a designer assign a withstand voltage to the snubber, the coupling capacitor and the bus capacitor one at a time.
• PD below 5 pC means the suppression parts do not become long-term insulation weaknesses inside the sealed ignition module.
• 此 HVCT8G screw-terminal and radial-lead packages tolerate the thermal cycling and vibration of a cutting cell while holding the tuned layout.
• 现场记录：在高压现场服务项目中，HVC 40 kV / 441 螺钉式电容器替代了国际一流部件，并且其结构与电气参数相匹配。 交叉参照 候选人。

交付和支持

标准部件的交货期为 4-8 周，配套组件的交货期为 6-10 周。1 kV 以上的设备可作为串联组件提供，并配有配套组件。 分级电阻器 that set the voltage division across the stack.

Send the measured L_σ, C_par and cable impedance, plus the worst-case ignition waveform. HVC returns snubber, clamping and coupling capacitor values with PD and ESR data for validation.

6. 常见问题

Q1: Where does the ignition spike come from?

Trapped magnetic energy, not the tank charge. When the gap quenches, the 235 mJ sitting in 10 µH of leakage inductance charges the switch-node capacitance and sets the peak voltage.

Q2: How large can the spike get?

It depends on node capacitance. V_spike = i_peak √(L_σ/C_par), so 217 A against 10 µH gives 30 kV at 520 pF and 68.6 kV at 100 pF. Shrinking the module usually shrinks the capacitance.

Q3: What damping ratio should the snubber target?

Between 0.5 and 0.8. A damping ratio of 0.5 gives 16 % overshoot, or 1.16× the steady value, and 0.8 gives 1.5 %. That keeps the first excursion near 1.2× the step.

Q4: How much of the spike is inductive kick?

With L in nanohenries and di/dt in kiloamperes per microsecond the answer comes out in volts. A 200 nH commutation loop with a 4.34 kA/µs fall adds 868 V, which is small next to the trapped-energy term.

Q5: Why terminate the torch cable?

An unmatched load sets Γ high and the reflected wave adds to the incident wave at the torch end, doubling or tripling the voltage. Matching Z_load to Z_c drives Γ to zero.

Q6: How is the insulation design verified?

Fire the igniter into a short, an open circuit and a marginal gap, log the switch and secondary voltages, and verify partial discharge at 120 % of the maximum stage voltage with a 5 pC limit on every component in the path.

7。 参考

1. E. Kuffel、WS Zaengl、J. Kuffel 高电压工程基础知识，第 2 版，Newnes，2000 年。
2. N. Mohan、TM Undeland、WP Robbins 电力电子学：变换器、应用和设计，第 3 版，Wiley，2003 年。
3. MIL-HDBK-217F 通知 2， 电子设备可靠性预测1995。
4. IEC 60384-1:2021， 电子设备用固定电容器 - 第 1 部分：通用规范，第 6 版。日内瓦：IEC，2021 年。
5. IEC 60664-1:2020， 低压供电系统设备绝缘配合——第1部分：原理、要求和测试，第 3 版。日内瓦：IEC，2020 年。
6. GB/T 7354-2018， 高电压试验技术局部放电测量。 北京: 中国标准出版社, 2018.
7. R. Sarathi 和 G. Koperundevi， 利用超高频技术研究直流电压下变压器油中单个导电颗粒的局部放电活动，IET 科学、测量与技术，第 3 卷，第 5 期，第 325-333 页，2009 年。https://doi.org/10.1049/iet-smt.2008.0136。
8. V. Belyaev， 真空绝缘体表面闪络模型，载于 ISDEIV 会议论文集。第 18 届真空放电和电绝缘国际研讨会，第 151-154 页，1998 年。https://doi.org/10.1109/deiv.1998.740596。
9. 富士IGBT应用手册，第5章——保护电路设计（REH984）富士电机，REH984，https://www.fujielectric.com/device/semi/technical/application/pdf/REH984/REH984_05.pdf。
10. 美国 6,760,215, 具有高电压击穿阈值的电容器DF DeVoe（与 Presidio 相关），2004 年授权，https://www.freepatentsonline.com/6760215.html。
11. HVC应用工程团队 高压陶瓷电容器 技术资源https://www.hv-caps.com