BATTERY CHARGER PULSE WIDTH MODULATION REGULATOR
Publication Date: 2003-Apr-02
The IP.com Prior Art Database
KEYWORDS: PULSE WIDTH MODULATION, BATTERY CHARGER, MULTIPLE INTEGRATION METHOD
Keywords: Pulse width modulation, battery charger, multiple integration method
To charge a battery from a wall adapter a discrete P-channel MOS or bipolar PNP transistor is used to switch the current from the wall adapter to the battery. There are different possible charge modes according to the nature and the state of the battery. One of these modes is the fast charge mode where huge currents (several hundreds of mA to more than 1A) are sent to the battery. In some cases this current can be a continuous one. In that case the gate of the discrete transistor is driven with a constant voltage, regulated through a feedback loop. In many other cases this current is pulsed (i.e., the transistor is switched on during only some part of the time, and it is closed the rest of the time). The two states alternate at a frequency called the switching frequency. It is generally desired in the case of a pulsed current to know the average value of the current provided to the battery and to be able to manage this value. In many cases a sense resistor (50 to 200 milli-ohm) is put in series between the discrete transistor and the battery. When the transistor is “on”, the voltage across the sense resistor is an image of the current loading the battery. A measurement of this voltage is made by an analog to digital converter (which can be very simple if the desired accuracy is small) and the result is sent to a microprocessor which will adjust the duty cycle of the switching frequency to get the desired result. One such technique is called Pulse Width Modulation (PWM). The accuracy of this method is limited because the precision of the converter is limited as the speed of the converter plus the reaction time of the microprocessor limit the speed of acquiring thin pulses to provide low currents. Moreover, this approach requires the use of a microprocessor; otherwise the above method is no longer feasible. Thus a new method and circuit is needed to provide low charging currents without using a microprocessor and with minimal waste while avoiding at least the above mentioned problems.
A new method (refer to Fig. 2) is to charge an internal capacitor (refer to Fig. 4) with a reference current during one period of the switching frequency, then to discharge it with an image of the current loading the battery. Thus, when the capacitor returns to its initial state the external PMOS transistor will be switched off (in the...