Battery Charger Constant Current Constant Voltage Design

A rechargeable lithium cell charges in two phases, and the circuit that does it has to switch between them without overshooting. The transition, the accuracy of the two set points and the way the charge is terminated are what separate a charger that lasts for years from one that shortens the life of every cell it touches.

The Two Phases of a Charge

During the first phase the charger holds the current at a fixed value, and the cell voltage rises as it accepts charge. That is the constant current phase, and its value is set from the cell capacity, usually at or below one times the capacity in amp hours.

When the cell voltage reaches its maximum, the charger switches to holding that voltage and the current begins to fall as the cell fills. The constant voltage phase is the longer one, and it ends when the current falls below a fraction of the set value, typically a tenth.

Both set points have to be accurate. A constant voltage that is fifty millivolts high reduces the cycle life measurably, and a current that is twenty percent high stresses the cell and the charger at the same time.

Sensing the Current and the Voltage

The current is measured across a resistor in series with the cell, and the sense amplifier has to work at the top of the cell voltage range where the common mode is close to the supply. A dedicated charger integrated circuit handles that requirement internally.

Kelvin connections at the sense resistor matter at these currents. A two amp charge through a ten milliohm resistor gives twenty millivolts, and a few milliohms of track or solder resistance is a visible error. The sense pair is taken from inside the pads, and the general practice is the same as for any low resistance measurement, described in our guide to mixed signal board design.

The cell voltage should be measured at the cell terminals rather than at the charger output. The trace and the connector between them carry the charge current, and their resistance adds a term proportional to current which makes the constant voltage phase end early.

The Transition Between Phases

The transition happens when the cell reaches its maximum voltage, and the loop has to hand over from current control to voltage control without a step. A slow handover produces a current overshoot that briefly exceeds the set value, and a badly compensated loop can oscillate.

Most integrated chargers implement both loops with a common error amplifier and a diode or a switch that selects whichever demand is lower. The behaviour at the corner is then smooth, because the two loops share the same compensation.

The charger should also detect a cell that will not accept charge. Where the voltage stays below a threshold for a set time, the cell is either deeply discharged or faulty, and a small pre charge current followed by a timeout protects against a shorted cell being charged at full rate.

charge termination and Top Up

charge termination ends the charge when the current falls below a fraction of the set value during the constant voltage phase. The threshold is usually between five and ten percent, and the charger then stops or switches to a maintenance mode.

A charger that simply stops leaves the cell to self discharge, and a product that is left plugged in will cycle between a partial charge and a recharge. A periodic top up that restores the cell when it falls by a few percent is gentler than a continuous trickle.

Timers are the safety net. A maximum charge time that is longer than the expected full charge but shorter than a fault condition ends the charge even if the current has not fallen. A separate watchdog on the charger active state catches a controller that has stopped communicating.

thermal regulation and Safety

thermal regulation reduces the charge current when the die of the charger or the cell becomes too hot. It is common in small products where the heat cannot escape, and the visible effect is a charge that takes longer in a warm room.

The cell temperature is measured with a thermistor bonded to the cell or built into the pack. Charging below freezing damages a lithium cell permanently, and charging above the upper limit is unsafe, so the temperature window is enforced in hardware rather than in software alone.

Protection is layered: the charger limits the current, the pack has its own protection circuit, and the system monitors both. A single failure should not be able to overcharge a cell, and the design description should state which layer covers which fault.

Layout and Thermal Design

Keep the power path short and wide between the input, the pass element and the cell connector. The charge current is continuous, so the voltage drop along a narrow trace is a real power loss and a real error in the measurement.

Place the sense resistor and its amplifier close together and keep the switching node of any step down charger as small as possible. The sense traces see a voltage of tens of millivolts, and a nearby switching node couples into them through the parasitic capacitance of the board.

Thermal relief for the pass element is part of the electrical design. The power dissipated depends on the difference between the input and the cell voltage, so a charger powered from five volts charging a nearly full cell dissipates most of the energy in the pass element rather than delivering it. The copper area under the device sets its temperature rise, and it should be sized from the worst case input voltage rather than the typical one.

Charging from a Shared Supply

A charger fed from a USB port has to respect the current the port will supply. A standard port offers five hundred milliamps, a dedicated charging port considerably more, and a charger that draws beyond the allowed figure can pull the supply down far enough to reset the device that is being charged.

The negotiation is handled by the controller, and the charger then limits its current to the agreed value. Where the port capability is unknown, the safe approach is to start at the low limit, watch the supply voltage and raise the current only while the supply holds up.

Where the same supply powers the system and charges the cell, the system has priority. The remaining current is offered to the charger, and the input current limit is set dynamically so that the supply never collapses under the combined load.

Verification and Faults

Verify the charger with an electronic load in place of the cell, set to draw the expected current, and with a voltage source to simulate the cell voltage. That lets the constant current and constant voltage set points be measured without cycling a real cell for hours.

Measure the current with a meter in the correct range and confirm the termination threshold by reducing the load current slowly. A charger that never terminates usually has a sense offset that keeps the measured current above the threshold.

A charger that runs hot and reduces its current is behaving correctly, and the fault is in the thermal design of the product rather than in the circuit. The release and inspection practices that keep these boards consistent are collected in our PCB design release checklist and judging PCB quality.

FAQ

Why does my battery take longer to charge in a warm room? That is the thermal regulation reducing the current. It is intended behaviour and protects both the charger and the cell.

How accurate does the constant voltage set point need to be? Within about one percent, which is fifty millivolts on a four point two volt cell. A high set point shortens the life of the cell.

What ends the charge? The current falling to a fraction of the set value in the constant voltage phase. A timer is added as a safety net in case that never happens.