DCDC电路的浪涌电流与防护

本文探讨了DC-DC转换器中出现的浪涌电流现象及其限制方法。介绍了浪涌电流产生的原因,包括两个主要峰值:第一个峰值发生在输入电压源开启时,第二个峰值发生在DC-DC转换器启动时。文章还详细讨论了如何通过电路设计来限制这些浪涌电流。

摘要生成于 C知道 ,由 DeepSeek-R1 满血版支持, 前往体验 >

Inrush Current in DC-DC Converters

September 22, 2008 Item MediaArticlesDesignEMC Test & Design GuideFilters / FerritesLightning & Surge

Some applications may require limiting the inrush spike into the input capacitors.

Steve Butler and Jeremy Ferrell, VPT, Inc., Blacksburg, Virginia, USA

BACKGROUND

Inrush current is the spike of current drawn by a power supply when it is turned on. A typical power system is shown in Figure 1. The input EMI filter will include some capacitance connected across the input line. The DC-DC converter will also include capacitance connected across its input and output. The load may include additional capacitance. Each of these capacitors requires current to charge them to their steady state voltage. This current is the inrush current.

Figure 1. Typical power system showing relevant capacitances.

Large inrush currents might be governed by a system level specification. There is a concern that high spike currents can create electromagnetic interference in adjacent circuitry, or trip an upstream circuit breaker or the over-current protection of a solid state power controller.

INRUSH CURRENT WAVEFORM

A typical inrush current waveform is shown in Figure 2. It has two current peaks. The first “inrush spike” peak current occurs when the input voltage source is turned on. This peak current flows into the EMI filter capacitors and the input capacitor of the DC-DC converter, charging them to their steady state value. The second current peak occurs when the DC-DC converter turns on. This peak current flows through the power transformer in the DC-DC converter to the output capacitor and into any load capacitance, charging them to their steady state value.

Figure 2. Typical inrush current waveform.

INRUSH SPIKE CURRENT

The first current peak is often referred to as the inrush spike. Its peak value and shape are highly dependent on the characteristics of the input source, specifically the voltage rise time and source impedance. A fast rising input voltage waveform, such as from a mechanical switch closure, will produce a very high and narrow current peak. A slower rising input waveform, such as from the output of some upstream electronics or a capacitor bank, will produce a lower, more controlled peak.

The peak value of the inrush current is determined by the equation, i=C*dv/dt. The capacitance is the total capacitance of the EMI filter and the input capacitance of the DC-DC converter. The dv/dt is the slope of the applied voltage waveform. This inrush spike is only an issue if the input voltage source has a very fast rise time. Normally the only thing that can produce a fast enough rise time is a mechanical switch or relay closure. If the source is a switching power converter, solid state power controller, or capacitor bank, the rise time will usually be much slower. Switching power converters usually have output voltage rise times on the order of a few milliseconds, solid state power controllers (SSPC) usually from 50us to 500us, and large capacitor banks usually cannot be charged in less than several milliseconds. These slow rise times will not produce excessively high current peaks. It is also important to evaluate not only the peak current, but the i2t of the current waveform to determine if an upstream fuse, circuit breaker or SSPC will trip due to inrush current.

TURN ON CURRENT

The second current peak of Figure 2 is also considered part of the inrush current. This peak occurs when the DC-DC converter turns on and draws current from the input to charge its output capacitance and any load capacitance. Typical turn-on current waveforms are shown in Figure 3. The turn-on current is the same whether the converter is turned on by applying the input voltage or via the inhibit signal.

Figure 3. Typical turn-on current waveform.

VPT’s DC-DC converters utilize a proprietary a magnetic feedback scheme with a well controlled internal start-up sequence and a precise output voltage soft start feature. The voltage soft start feature ensures the output ramps up in a controlled manner, with limited dv/dt. Due to the soft start, the input current usually does not exceed the steady state input current of the converter during turn-on, see the normal curve shown in Figure 3.

VPT’s DC-DC converters also feature continuous constant output current limit. They will supply full rated current into any load, they do not hiccup or shut down and restart. This allows them to start any load capacitor, regardless of size. For cases with very large load capacitance, the DC-DC converter might enter this constant current limit mode during turn-on. In this case the input current would not exceed approximately 1.5 times the steady state input current, see the large load capacitance curve in Figure 3. This is not high enough to cause any interference or trip upstream protection devices. For VPT’s DC-DC converters, this second inrush peak will not cause adverse effects in the system design.

ACTIVE INRUSH LIMITING

In some applications there is a requirement to limit the inrush spike current into the input capacitors. The only way to do this is to insert a series element into the circuit in front of those capacitors. Figure 4 shows a very basic inrush limiting circuit. The series resistor R1 will limit the input current until the input capacitors are charged. The relay S1 will then close to allow the full current to flow to the downstream DC-DC converter. The relay coil can be driven from the 28V input such that it will be somewhat automatically controlled.

Figure 4. Discrete active inrush current limiting circuit using a series resistor.

An inductor could also be used to limit inrush current. It would not need to be bypassed since it can carry DC current with low loss. However a large inductance value is usually required to limit the inrush sufficiently. Care must be taken since this inductor can ring with the input filter or can interact with the DC-DC converter’s internal feedback loop, causing system instability. Additional components are usually necessary to damp the resulting resonance.

Another practical circuit is shown in Figure 5. This circuit uses a series MOSFET Q1 placed in the negative power lead. Q1 is normally off, with its gate pulled low through R2. When input voltage is applied, the gate will charge through R1. The charge time and turn-on of Q1 will be slowed by C1. R1 and C1 can be chosen to allow the input capacitors to charge slowly, limiting the inrush current. After the input capacitors are charged, the gate of Q1 will charge until it is limited by the zener and Q1 will remain fully on.

Figure 5. Discrete active inrush current limiting circuit using a series MOSFET.

This circuit can be modified with Q1 located in the positive lead. It can be driven in the same manner by using a P-channel MOSFET. Or it can use an N-channel MOSFET but with the gate driven by a charge pump or isolated supply. Various other active inrush limiting circuits exist. They all use some series device in the input power leads and act in approximately the same way. It is always important that once the input capacitors are charged, the series device is bypassed or turned fully on to limit line impedance and power dissipation. It is also important that the active inrush control not introduce any noise into the input line since it is located upstream of the EMI filter.

INPUT MODULES WITH INRUSH LIMITING

Several of VPT’s input modules include built-in inrush current limiting. They are listed in Table 1. These modules each use a series N-channel MOSFET in the positive lead. The N-channel MOSFET achieves the lowest ON-state resistance to minimize power losses. Locating it in the positive lead leaves the return path unbroken, simplifying system design. And, this series MOSFET is used for a dual purpose in these modules; it also provides input voltage transient protection.

The DV-704A or DVMN28 contain both an EMI filter and inrush limiting. The two circuits are optimized to work together. The inrush circuit limits any current flowing into the EMI capacitors, but does not introduce any additional EMI into the input lines, as is possible with a discrete circuitry. The VPTPCM-12 contains inrush limiting circuitry which will limit inrush current into its capacitors and those of downstream DC-DC converters. But it also contains switching elements, so it may require additional unprotected EMI filtering at its input.

CONCLUSION

Inrush current is the peak current which flows into the power converter when input voltage is applied or at turn-on. Most system inrush current specifications will be met with VPT’s DC-DC converters alone. Some applications may require limiting the inrush spike into the input capacitors. This will require additional circuitry, or one of VPT’s input modules with built-in inrush limiting.

ABOUT THE AUTHOR

Steve Butler is Vice President of Engineering and Jeremy Ferrell is a Senior Design Engineer at VPT, Inc. in Blacksburg, Virginia. VPT is a vendor of DC-DC converter and EMI filter modules and can be reached ad www.vpt-inc.com.

<think>好的,我现在需要帮助用户解决如何抑制DCDC前端的浪涌电流的问题。用户之前已经询问过关于限流电路的设计,现在他们可能在实际应用中遇到了浪涌电流的问题,或者想进一步优化他们的设计。 首先,我需要回顾之前的对话,确保这次回答保持一致性。用户之前的问题是关于使用两个PNP晶体管组成的限流电路,限流300mA,并且负载是DCDC转换器。之前的回答中提到了浪涌电流处理,比如并联电容和使用RC延时电路。现在用户特别关注如何抑制浪涌电流,可能需要更详细的方案。 接下来,我需要考虑用户可能的身份和需求。他们可能是电子工程师或爱好者,设计电源电路时遇到了启动时的浪涌电流问题。浪涌电流可能导致限流电路误触发,或者损坏元件,影响DCDC的正常工作。用户需要多种方法来抑制这种浪涌电流,可能需要从简单到复杂的解决方案,以及实际应用中的注意事项。 我需要先解释浪涌电流的产生原因,这样用户能更好地理解解决方案。DCDC输入端通常有大容量电容,在启动瞬间电容充电会导致瞬间大电流。此外,输入电压突然加上时,电感等元件也可能导致电流尖峰。 然后,我需要列举常见的抑制方法,每种方法的原理、优缺点以及适用场景。比如使用NTC热敏电阻、缓启动电路、限流电阻+继电器、有源限流电路、保险丝的选择、输入电感、TVS二极管等。需要详细说明每个方案的工作原理,可能的设计参数,比如NTC的选型,缓启动电路的RC时间常数计算等。 同时,要结合用户之前的电路,比如他们使用了PNP限流电路,可能需要补充其他措施,比如并联电容或增加TVS二极管。还要提到设计时的验证要点,比如测试浪涌电流的波形,元件的温升等,确保方案的可靠性。 需要确保回答结构清晰,分点列出,每个方法有标题、原理描述、具体实现和注意事项。使用用户熟悉的术语,比如之前提到的RC延时、电解电容等,保持一致性。同时,避免过于技术化的术语,确保易懂。 最后,总结关键点,帮助用户根据实际需求选择合适的方案。可能用户需要成本低但效率不高的方案,或者高性能但复杂的方案,所以需要提供不同选项供其选择。</think>以下是抑制DCDC前端浪涌电流的常见方法及技术细节分析: --- ### **一、浪涌电流成因** DCDC输入端通常有**大容量滤波电容**,上电瞬间电容充电表现为短路状态,导致: $$ I_{inrush} = \frac{V_{in}}{ESR_{电容} + 线路阻抗} $$ 典型值可达数十安培,可能触发保护电路或损坏元件。 --- ### **二、抑制方案及实施** #### **1. NTC热敏电阻** - **原理**:利用负温度系数特性 - 冷态高阻值限制电流 → 发热后阻值下降 - **选型要点**: $$ R_{25℃} > \frac{V_{in}}{I_{max}} \quad (例如12V/3A系统选5D-9型,5Ω/3A) $$ - **优点**:简单、低成本 - **缺点**: - 连续工作会发热,需留散热间距 - 频繁开关机时冷却不足导致失效 #### **2. 缓启动电路(Soft-Start)** - **电路拓扑**: ```text MOSFET + RC延时 12V → MOSFET漏极 → 源极→ DCDC │ R(10kΩ) │ C(10-100μF) GND ``` - **工作过程**: - 上电时电容充电,MOSFET栅极电压缓慢上升(时间常数$τ=RC$) - MOSFET从线性区逐渐进入饱和,实现**电压斜坡上升** - **关键参数**: $$ t_{rise} ≈ 3RC \quad (例如R=10kΩ, C=47μF ⇒ t≈1.4s) $$ #### **3. 限流电阻+继电器** - **两级结构**: 1. 启动阶段:电流通过限流电阻(如2Ω/10W) 2. 电容充电完成后,继电器短路电阻 - **优化变种**: - 用MOSFET替代继电器,通过比较器检测电压触发 #### **4. 有源限流电路** - **方案示例**: ```text ┌─ 运放控制MOSFET电流 │ 检测电阻Rs(0.1Ω) → 运放(如LM358) → 驱动MOSFET └─ 可设定精确限流阈值(如I_limit=5A) ``` - **特点**: - 响应速度快(μs级) - 支持动态调整,但成本较高 #### **5. 输入保险丝选择** - **慢熔型保险丝**(如Littelfuse 215系列): - 可承受5~10倍额定电流持续100ms - 选型公式: $$ I_{rated} ≥ 1.5 \times I_{steady} $$ #### **6. 输入串联电感** - **铁氧体磁珠**(如Murata BLM系列): - 抑制高频浪涌分量 - 参数选择: $$ Z@100MHz > 100Ω \quad (例如BLM18AG102SN1) $$ #### **7. TVS二极管保护** - **双向TVS**(如SMBJ15CA): - 吸收电压尖峰,防止浪涌引发过压 - 布局要点:尽量靠近DCDC输入端 --- ### **三、组合方案示例** ```text [推荐高可靠性设计] 12V输入 → NTC(5D-9) → TVS(SMBJ15CA) → 缓启动电路 ↓ 10μH电感 ↓ 470μF电解电容 ↓ DCDC模块 ``` --- ### **四、设计验证要点** 1. **示波器测试**: - 测量上电瞬间电流波形,确认峰值<目标值 - 检查电压上升时间是否符合预期(一般建议50-200ms) 2. **热测试**: - NTC在连续工作1小时后表面温度应<85℃ 3. **重复开关测试**: - 快速开关机100次,观察保护电路是否失效 --- ### **五、选型对比表** | 方案 | 成本 | 效率 | 复杂度 | 适用场景 | |--------------|-------|------|--------|------------------------| | NTC | $ | 中 | 低 | 低成本、非频繁开关 | | 缓启动电路 | $$ | 高 | 中 | 中功率系统(10-100W) | | 有源限流 | $$$ | 高 | 高 | 精密仪器、工业设备 | | 继电器方案 | $$ | 中 | 中 | 大功率系统(>200W) | --- 根据系统功率等级、成本预算及可靠性要求选择合适方案。对于**12V/300mA级系统**,建议优先采用 **NTC+缓启动电路组合**,兼顾成本性能。
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值