Protected Battery Startup Measurements带保护电池的启动测量

Connecting the Claude Chime power board to its protected battery pack dropped the measured pack output from 4.03 V to 1.49 through 1.57 V. Yet a 3.8 V bench supply with a 50 mA limit could start both unloaded boost rails on the same board. Together, those observations raise questions about how each source enters operation and which part of the circuit an instrument is measuring.

Claude Chime 电源板接入带保护电池时，我在电池包输出端量到的电压从 4.03 V 降到了 1.49 至 1.57 V。同一块板却能用 3.8 V、限流 50 mA 的台式电源启动两路空载升压。这两个结果放在一起，最值得继续研究的是电源怎样进入工作状态，以及仪器量到的究竟是哪一段电路。

Cell and Pack-Output Voltage Measurements电芯与电池包输出电压测量

A protected pack contains a cell, protection circuitry and a series switching path. A meter across its output terminals measures the voltage available to the load. If the protection switch opens, the load, leakage and internal paths can leave a low output reading while the cell still holds charge. Before interpreting a reading near 1.5 V as a cell condition, I need to identify the measurement nodes.

带保护的电池包包含电芯、保护芯片和串联开关路径。万用表接在包的输出正负端，读到的是负载能够使用的端电压。保护开关断开时，即使电芯仍有电，输出端也可能因负载、漏电及内部路径呈现很低的读数。看到 1.5 V 左右，必须先确认测量节点，再讨论电芯状态。

The behavior made me suspect startup activity in the pack's DW01A and 8205A protection path. Fortune's DW01A documentation describes overcurrent detection using the CS-pin voltage and a corresponding delay. Converting that voltage threshold into a current requires the impedance of the monitored path. The threshold alone is no universal current rating for every protection board.

这组现象让我怀疑电池包上的 DW01A 与 8205A 保护路径在启动时发生了动作。DW01A 原厂资料说明，电流检测使用 CS 引脚的电压，并结合相应延时判断过流。门限是检测电压，换算为电流还要知道受监测路径的阻抗。不能把一个电压门限直接当成所有保护板通用的电流额定值。

An ideal capture for separating cell droop from an opened protection path would include cell voltage, pack output and connection current together. Measurements across different battery references require an appropriate differential arrangement, so common-ground oscilloscope probes do not accidentally bypass the protection path. The measuring connection itself can change the circuit, especially with low-side protection.

要区分电芯自身下跌与保护路径断开，理想记录应当同时包含电芯端电压、包输出电压和接入电流。测电池不同参考点时，要使用合适的差分测量方式，避免示波器共地探头把保护路径意外短接。测量连接本身会改变电路，这一点在低侧保护结构里尤其需要留心。

Capacitor Charging and Startup Current电容充电与启动电流

At connection, capacitors need charge. TI's inrush-current application note expresses that current as I = C × dV/dt: for the same capacitance, a faster voltage rise demands more current. Added capacitance that helps with operating load transients also increases the charge needed at startup.

电源刚接入时，电容需要获得电荷。TI 的浪涌电流说明用 I = C × dV/dt 描述这部分电流。相同电容量，电压建立得越快，所需电流就越大。为了帮助运行中的负载瞬态而增加电容，也会增加接电时需要补充的电荷。

The schematic shows three 22 μF capacitors across the boost inputs, 4.7 μF at the charger's battery terminal and two 100 nF capacitors, all directly connected to BAT+. Their nominal sum is about 70.9 μF. For an illustrative linear 4 V rise in 100 μs, those capacitors alone require about 2.84 A; extending the rise to 1 ms reduces that to about 0.284 A. This is an ideal estimate. Effective capacitance, wiring impedance and source response alter the actual waveform.

原理图里直接接在 BAT+ 上的电容，包括两路升压输入的 22 μF、22 μF、22 μF，充电芯片电池端的 4.7 μF，以及两颗 100 nF。只按标称值相加，已经约有 70.9 μF。取 4 V 电压变化、100 μs 线性上升作示例，仅这部分电容就对应约 2.84 A 的充电电流；把上升时间延长到 1 ms，则约为 0.284 A。这是理想估算，实际电容量、连线阻抗和电源响应都会改变波形。

Output capacitors need examination along their own paths. The TLV61048 datasheet describes internal soft start, but the external inductor and rectifier also provide a path that can charge the output when power is first connected. Current can flow before the internal reference begins its controlled ramp. Direct input charging, output precharge and regulated boost startup should therefore be distinguished when examining the input current.

输出电容还要沿各自路径继续计算。TLV61048 手册描述了内部软启动，但这颗芯片使用外部电感和整流二极管，接电时可以先沿电感与二极管向输出充电。芯片开始调节内部参考之前，已经可能有电流流过。因此，检查启动电流时，需要把直接输入充电、输出预充及进入升压调节后的过程分开。

Supply Current Limiting and Battery Protection电源限流与电池保护

Successful startup at a 50 mA limit is useful, but the bench supply has a different response from the protected pack. In current limit, its voltage may rise more slowly. Output capacitance, control response and connection details also affect the initial current. A protected battery may instead interrupt discharge once its detection conditions are met. The same board can consequently follow different startup trajectories with the two sources.

50 mA 限流启动成功很有价值，但它对应的是另一种电源响应。台式电源进入限流时，输出电压可能上升得更慢；输出电容、控制响应和连接方式也会影响最初那一小段电流。带保护电池则可能在检测条件满足后切断放电路径。两者面对同一块板，可以走出不同的启动过程。

Resistance mode helps look for a persistent low-resistance path, but does not reproduce power-on charging. Its test current and voltage differ from the battery connection, potentially putting semiconductor devices in different states. I would use resistance checks to narrow the persistent-short possibility, then voltage and current captures to investigate connection behavior. Each measurement has a different job.

万用表电阻挡适合先查持续低阻，却无法重放上电时的充电过程。它的测试电压和电流与真正接入电池不同，半导体的状态也可能不同。我会把电阻检查用于缩小持续短路的可能范围，再用电压、电流波形研究接入瞬间，各种测量回答各自的问题。

At the time, disconnecting the board restored pack output; applying an external supply at the pack output was also followed by a successful battery start. That makes the initial state of each trial worth controlling. Residual capacitor charge, protection recovery and whether the board remained connected can all affect repeatability. I would fix those conditions before comparing further trials.

当时断开主板后，包输出电压会恢复；给包输出端接过外部电源后，下一次启动也成功过。这让我更关心每次试验开始前的状态。电容有没有放完电、保护是否已经恢复、主板有没有持续连接，都会影响可重复性。继续试验时，应先固定这些条件，再比较结果。

Startup Waveform Measurements启动波形测量

Triggering on battery connection would capture pack-output voltage and current together. A current probe is preferable; a shunt requires accounting for its additional voltage drop and reading it differentially. A peak alone is insufficient. Its duration, the beginning of output collapse and whether current is then interrupted belong in the same capture.

以电池接入事件触发示波器，可以同时记录电池包的输出电压和电流。电流测量优先考虑电流探头；若使用分流电阻，则需计算它额外引入的压降，并用差分方式读取。只量到一个尖峰还不够，还要看它持续多久、输出何时开始下降，以及下降之后电流是否被切断。

If current rises first, pack output suddenly falls and cell voltage remains relatively high, the protection path deserves priority. If the cell voltage also falls substantially, cell resistance and connection drops need investigation. The waveform can then guide a choice of precharge, controlled rise time or a revised startup sequence. Both converters share the connection event, so their combined demand needs observing at the common input.

如果电流先增大，随后包输出突然降低，而电芯端仍保持较高电压，保护路径就值得优先检查。如果电芯端也明显下跌，则要继续查电芯内阻和连接压降。根据波形，再选择预充、受控上升斜率或调整启动安排，改动会更有针对性。两颗 Boost 共用同一次接电事件，合计电流也应在共同输入端观察。

Those synchronized measurements provide a concrete continuation for the fault investigation. Whether current rises before voltage falls, and which nodes begin to separate, can progressively narrow the cause. Once startup is visible at that level, circuit changes have a much clearer basis.

启动掉压的后续排查，可以从这组同步测量继续。电流先增大还是电压先下降、哪两个节点开始分离，这些细节能够逐渐缩小原因范围。把启动过程测清楚以后，电路该改哪里，也就有了具体依据。