Many portable products require some form of battery-management system. In large systems, the battery-management system may have to deal with different battery chemistries and cell configurations.
Today’s system designers often turn to intelligent battery charging as part of the power-management strategy to deal with these battery-pack variations. At the same time, some users want to continue using traditional, low-cost, “dumb” battery packs. This is particularly true in lowvolume products that operate in remote sites with unreliable ac power. Since the duration of power loss varies from site to site, different watt-hour-rated battery packs with different cell arrangements are employed.
To implement an intelligent battery-charging strategy in a “handsoff” mode of operation requires the use of a microcontroller. The circuit shown in the figure interfaces to a microcontroller and provides user-definable voltage and current charge ranges. To charge a wide variety of battery chemistries and cell configurations, the typical ranges for this circuit are 2 to 22 V for charging voltage and 0 to 3 A for charging current.
The circuit uses an LT1446 dual 12-bit DAC to provide the microcontroller-to-analog interface. Twelve bits of control provide high-resolution steps equal to 0.02% of the fullscale control range. The DAC has an on-board voltage reference with ±0.7% accuracy. This reference, combined with 0.5% resistors, is more than accurate enough for constantvoltage charging of Li-Ion batteries.
The DAC’s outputs are connected to an LT1511 battery-charger IC by an LT1490 dual, low-power, rail-to-rail op amp. A1 acts as a voltage error amp, comparing the reference voltage created by the DAC output with the divided-down feedback voltage coming from the charger’s output. By using integer divisors, such as 2, 3, 4, or 5, you can easily correlate the DACprogrammed output voltage with the battery-charging output voltage.
The output of the voltage-error amplifier, A1, is connected to the OVP pin of the LT1511 for voltage control. When the battery charger is in constant-voltage mode, the voltage on the LT1511 pin will be around 2.465 V. If the charger is in constant-current mode, the voltage will be 0 V.
Resistor RF and capacitor CF provide a gentle rolloff of amplifier A1’s ac gain. Amplifier A2, combined with transistor Q1 and resistor R5, is configured as a programmable currentsink that’s connected to the PROG pin of the LT1511 through an RC compensation circuit. Transistor Q1 sinks current that’s directly proportional to the charging current flowing in the battery. Resistors R3 and R4 scale the DAC output voltage by half to be below the LT1511 PROG pin’s 2.465-V reference-voltage control range.
The actual charging power is generated by the LT1511, a PWM-based battery-charger IC using Linear Technology’s patented “input current regulation” feature. This feature prevents overloading of the ac adapter. Concurrently, it allows the fastest possible battery charging at any given time, regardless of variations in power drain by the host system. Note that the boost pin of the LT1511 is connected to the 5-V supply. This allows for high-voltage operation, lower dropout voltage, and increased operating efficiencies.
Design calculations are as follows: the undervoltage lockout threshold is 6.7 V at the LT1511 UVIN pin. The VIN turn-on threshold is VIN = \[5k × (VIN − 6.7)/6.7 V\]. When the input voltage drops below the minimum threshold, the LT1511 will shut itself down. Also, the 5-V regulator will be shut down via its UVOUT pin to help conserve power. The current limit functions for both RS1 and RIN are referenced against the LT1511’s internal 100-mV references, such that R = 100 mV/ ILIMIT.
Next, determine the maximum output voltage to be used for charging. R1 sets the upper voltage range limit. The value is R1 = \[162k × (VBATMAX − 4.095 V)/4.095 V\]. Because the maximum programmable batterycharging current is a function of the current pulled out of the LT1511 PROG pin to ground, R5 must match the maximum current limit value set by RS1. With RS2 = RS3 = 200 Ω, R5 (200 × 2.465 V)/(IBATMAX × RS1). R5 may need to be adjusted down a little to compensate for errors caused by the base current of Q1 and internal LT1511 errors.
A design example using a 3-A, 28-V power source yields the component values shown in the table.