Dario Nurzad (Application Engineer, National Semiconductor Corporation)
White LEDs are typically driven by a constant DC current source to maintain a constant brightness. In portable applications powered by a single Li-Ion battery, the sum of the voltage drops across the white LED and the current source can be higher or lower than the battery voltage, which means that the white LED needs to boost the battery voltage at some point. The best way to accomplish this is to use a step-up DC-DC converter, which greatly optimizes efficiency, but at the cost of increased cost and PCB area. Another way to boost the battery voltage is to use a charge pump, also known as a switched capacitor converter. This article will examine in detail how this component works.
The basic principle of charge pump
A capacitor is a device that stores charge or electrical energy and discharges at a predetermined rate and time. If an ideal capacitor is charged with the ideal voltage source VG (see Figure 1a), the charge is immediately stored according to the Dirac current pulse function (Figure 1b). The total amount of charge stored is calculated as follows: Q = CVG
The actual capacitor has an equivalent series impedance (ESR) and equivalent series inductance (ESL), neither of which affects the ability of the capacitor to store electrical energy. However, they have a large impact on the overall conversion efficiency of the switched capacitor voltage converter. The equivalent circuit for actual capacitor charging is shown in Figure 1c, where RSW is the resistance of the switch. The charging current path has a serial inductor that can be reduced by proper component layout.
Once the circuit is powered up, transient conditions of exponential characteristics are generated until a steady state condition is reached. The parasitic effects of the capacitor limit the peak charging current and increase the charge transfer time. Therefore, the charge accumulation of the capacitor cannot be completed immediately, which means that the initial voltage change across the capacitor is zero. The charge pump utilizes this capacitive characteristic, as shown in Figure 2a.
Figure 1. Charging a capacitor from a voltage source (Figures a and b are ideal and c and d are actual).
The voltage conversion is implemented in two phases. During the first phase, switches S1 and S2 are closed, while switches S3 and S4 are open and charged to the input voltage:
VC1+ VC1- = VC1+ = VIN
VC1+ ─ VC1- = VOUT ─ VIN = VIN → VOUT = 2VIN
In the second phase, switches S3 and S4 are closed and S1 and S2 are open. Because the voltage drop across the capacitor does not change immediately, the output voltage mutates to twice the input voltage value. This method can be used to achieve voltage doubler. The duty cycle of the switching signal is typically 50%, which usually produces the best charge transfer efficiency. Let's take a closer look at how the charge transfer process and the switching capacitor converter parasitics affect its operation.
Figure 2..a. Charge pump circuit, b. Correlation waveform.
The steady-state current and voltage waveforms of the switched capacitor voltage doubler are shown in Figure 2b. According to the principle of power conservation, the average input current is twice the output current. In the first phase, the charging current flows into C1. The initial value of this charging current is determined by the initial voltage across capacitor C1, the ESR of C1, and the resistance of the switch. After C1 is charged, the charging current decreases exponentially. The charging time constant is several times the switching period, and a smaller charging time constant will result in an increase in peak current. During this time, the output capacitor Chold provides the amount of linear discharge of the load current, and the discharge amount is equal to:
In the second phase, C1+ is connected to the output, and the discharge current (the current is the same as the previous charging current) flows through C1 to the load. At this stage, the output capacitor current changes by approximately 2IOUT. Although this current change should produce an output voltage change of 2 Iout ESR C hold, using a low ESR ceramic capacitor makes this change negligible. At this point, CHOLD is charged at the linear potential of the following battery:
When C1 is connected between the input and ground, C hold discharges according to the following linear potential:
The following is the total number of output crest peak-to-peak voltage values:
Higher switching frequencies can use smaller output capacitors to achieve the same chopping. The parasitic effect of the charge pump causes the output voltage to decrease as the load current increases. In fact, there is always 2IOUT of RMS current flowing through C1 and two switches (2Rsw), resulting in the following power dissipation:
In addition to these pure resistive losses, the RMS current of IOUT flows through the equivalent resistance of the switched capacitor C1, resulting in a power dissipation of:
The RMS current flowing through CHOLD is equal to IOUT, resulting in the following power dissipation:
All of these losses can be summarized by the following equivalent output resistance:
In this way, the output voltage of the charge pump can be imitated by the following equation:
VOUT = 2VIN ─ Iout Rout
In summary, because of the low ESR of the ceramic capacitor and the high switching frequency, the output ripple and the output voltage drop depend on the switching resistance. Additional voltage conversion can be achieved with more switches and capacitors. Figure 3 shows the circuit using this characteristic of the capacitor. Similarly, voltage conversion is done in two phases. In the first phase, switches S1 through S3 are closed and switches S4 through S8 are open. So C1 and C2 are connected in parallel, assuming C1 is equal to C2, then charging to half the input voltage..
The output capacitor CHOLD provides the output load current. As this capacitor discharges, the output voltage drops below the desired output voltage, and the second phase is activated to increase the output voltage above this value. In the second phase, C1 and C2 are connected in parallel and connected between VIN and VOUT. Switches S4 to S7 are closed, and S1 to S3 and S8 are open. Because the voltage drop across the capacitor does not change, the output voltage jumps to 1.5 times the input voltage..
Figure 3. Switched-capacitor circuit with 1x and 1.5x gain.
The voltage boost is done through the following modes: By turning off S8 and keeping S1 to S7 open, the voltage conversion can achieve 1x gain.
Pulse frequency modulation (PFM) scheme
Figure 4 illustrates a simplified PFM voltage regulation scheme that utilizes many gains. The down-regulated output voltage is compared to a 1.2V voltage reference through a PUMP/SKIP comparator. The PUMP/SKIP comparator output voltage rises linearly at startup, providing a soft-start function. When the output voltage exceeds the desired limit, the component will not turn on and the power supply current consumed will be small. During this idle state, the output capacitor provides an output load current. As this capacitor continues to discharge and the output voltage drops below the desired output voltage, the charge pump is activated until the output voltage again rises above this value.
The main advantages of the PFM conditioning architecture at light loads are obvious. The load energy is usually supplied through the output capacitor. The supply current is very low and the output capacitor only needs to be recharged through the charge pump occasionally.
In summary, the regulated charge pump cannot maintain high efficiency over a wide input range because the input-to-output current ratio is adjusted according to the basic voltage transition, any value that is lower than the input voltage multiplied by the charge pump gain. The output voltage will result in additional power dissipation within the converter and the efficiency will decrease proportionally.
The ability of the converter to vary the gain based on the input/output ratio allows for the best efficiency over the entire input voltage range. Ideally, the gain should be linear. In reality, given the fixed number of capacitors and switches, it is only possible to achieve a limited gain configuration.
In Figure 4, the input voltage is adjusted and fed into the forward junction of the three comparators. All reverse junctions of the comparator are connected to the output voltage. Depending on the input-to-output voltage ratio, the output of the comparator provides a gain control circuit with a 3-bit word that is used to select the minimum gain G so that the desired voltage transition can be achieved. However, in white LED applications, choosing the correct gain G is not just based on the input and output voltages.
Figure 4. Block diagram of the switched capacitor voltage regulator.
Highly integrated charge pump dual display LED driver
Taking NS's LM27965 charge-pull dual-display LED driver as an example, the D1A-5A or D1B-D3B outputs can be connected together to drive one or two LEDs at a higher current. In such a configuration, all five parallel current outputs can drive one LED. The LED current for the D1A-5A should be selected so that each output current can be set to 20% of the desired total LED current. For example, if 60 mA is the desired single LED drive current, the appropriate RSET should be chosen such that the input current through each current sink is 12 mA. The available diode output current, maximum diode voltage, and all other parameters provided in the electrical parameter table are identical to the standard 5-LED application circuit.
At higher input voltages, the LM27965 operates in Pass-Mode, allowing the output voltage to track the input voltage. As the input voltage continues to decrease, the voltage on the Dxx pin also drops (VDXX = VPOUT VLEDx). Once any activated Dxx pin reaches a voltage close to 175mV, the voltage pump will switch to a gain of 3/2x. This switching ensures that the current flowing through the LED is not affected by the lack of sufficient voltage margin across the LED. The first and second sets of outputs utilize on-chip LED forward voltage detection on each Dxx pin to optimize charge pump gain for maximum efficiency. Due to the nature of the detection circuit, it is not recommended to suspend any DxA (D1A-D4A) or DxB (D1B-D2B) pins if any of the LED groups will be used during normal operation. If the DxA and / or DxB pins are left floating, the charge pump will be forced into the 3/2x mode over the entire VIN range.
If the D5A is not in use, it is recommended to ground the driver pins and set the EN5A bit of the general-purpose buffer to 0 to ensure proper gain conversion. With a universal buffer, the D3B drive can be fully turned on or off during work. Activate the diode monitoring circuit and disable the driver. If D3B is not used, it is recommended to ground the driver pins and set the EN3B bit of the general-purpose buffer to 0 to ensure proper gain conversion.
Figure 5. Typical Application Circuit for LM27965
in conclusion
The use of switched capacitors has certain advantages over inductive based switching methods. One obvious advantage is the elimination of inductance and associated electromagnetic design issues. Switched-capacitor converters typically have relatively low noise and minimal radiated EMI. In addition, the application circuit is simple and requires only a few small capacitors. Because in the absence of inductance, the final PCB assembly height is typically smaller than equivalent switching converters.
Ultra Thin Battery,Ultra Slim Battery Pack,Thin Lithium Ion Battery,Ultra Light Optics Battery,Polymer lithium battery,Cylindrical Lithium Battery
Langrui Energy (Shenzhen) Co.,Ltd , https://www.langruibattery.com