如何最大限度地减少功耗锂离子线性充电器-How to Min

Abstract: Techniques are described for minimizing power dissipaTIon in linear battery chargers. Beginning with a stable wall-cube switching power source, methods are described to limit the dissipaTIon in the linear charging circuit. Circuits are provided, calculaTIons are shown, heat sinking for the PMOS pass transistor is discussed, and suitable pass transistors are suggested.

IntroducTIonData sheets for single-cell Li+ linear chargers seldom discuss power dissipation or how to deal with heat dissipation. High input voltage and charge current increase the amount of power the pass element must handle. This application note discusses how to maximize charging current while maintaining safe device and system temperature limits.

Use a Proper DC Input SourceA low voltage input reduces the power dissipation. In order to charge the single-cell Li+ battery, we need a well regulated 4.2V±1% or 4.1V±1% (depending on battery chemistry) output. The input voltage needs to be higher to cover the voltage drops between the battery positive terminal and the input DC source. Figure 1 shows these for a typical charger.


Figure 1. Voltage drop contribution.

Vin = Vsense + Vpmos + Vtrace + Vdiode + 4.2V

The minimum input can be described as below.

Vin(min) = Rsense × Icharge + Rds(on) × Icharge + Rtrace × Icharge + (Vthmax(d) + Rd × Icharge) + 4.2V

Where Vdiode = Vthmax(d) + Rd × Icharge, Vthmax(d); Diode turn on threshold voltage, Rd; Diode series resistance

As we see in the above equation, the charger requires higher input voltage if the charge current (Icharge) is increased. Below is actual data from an example circuit (Figure 4) when the charge current is 500mA.

Vin = 0.303V(Vdiode) + 0.060V(Vsense) + 0.112V(Vpmos) + 0.000V(Vtrace) + 4.2V

Vin = 4.68V

Schottky diode: Zetex ZHCS1000,

PMOS FET: Fairchild FDC636P,

Rsense = 105mΩ,

Rtrace: 40mils wide, 0.5" long, and 1 oz copper trace. This value depends upon PCB layout and battery contacts.

Since this data was taken from one prototype, we should also consider the tolerances of each parameter. A 5V±5% well-regulated switching mode AC adapter will provide some margin, to account for tolerances. The AC adapter does not need an accurate current limit since the charger has a current control, but the AC adapter maximum current capability must be 200-300mA higher than the fast charge current of the linear charger. Figure 2 shows an example of an AC adapter using a MAX5021 low-power, current-mode PWM controller.


Figure 2. 5V/1A AC adapter.

Optimize Charge Current and Power DissipationFigure 3 shows the circuit used for testing. It is a linear charger with a 500mA charge current and a 6 hours timer limit.


Figure 3. The MAX1898 single-cell Li+ linear charger.

Total power dissipation for the linear charger can be expressed as below.

Pdiss = (Vin - Vbatt) × Icharge

To decide the fast charging current, we need to calculate the worst-case allowable power dissipation on the P-MOSFET Q1.

Power dissipation on the Q1 is expressed as below:

Pdiss(Q1) = Vds(Q1) × Icharge

Vds(Q1) = 5V - VD1 - Icharge × Rcs - Vbatt

Where VD1: D1 forward voltage drop, Rcs: internal current sensing resistor.

Also, the junction temperature of the P-MOSFET should not exceed its maximum limit = 150°C at any operating conditions.

Tj = Ta + RΘJA × Pdiss(Q1)

Table 1 shows some possible P-MOSFET products that can be used in the charger. Even though the specifications show quite high maximum power dissipations, we should be cautious of the PCB mount condition. The "1 in_ pad of 2oz Cu on FR-4 board" specified for package rating on many MOSFET devices may not be realistic for many applications. Instead, the following design procedure yields more practical results.

Table 1.
  Package