Step-up DC/DC voltage converter with improved transient current capability

Boost DC/DC voltage converter with improved transient current performance

The present invention relates to the design, operation and performance of switching power supplies used in DC/DC conversion and voltage regulation, and relates to semiconductor devices used in these converters. In particular, the present invention focuses on step-up DC/DC conversion, that is, where the output voltage exceeds the input voltage, and particularly focuses on the DC/DC conversion, where the output voltage is significantly greater than the minimum input voltage.

This application is a partial continuation of Application No. 11/890,818 and Application No. 11/890,956 each filed on August 8, 2007 and incorporated by reference in this article.

Voltage regulation is usually required to prevent the supply of power for various microelectronic components (such as digital ICs, semiconductor memory, display modules, hard disk drives, radio frequency (RF) circuits, microprocessors, digital signal processors, and analog ICs) Voltage changes, especially battery-powered applications (such as cellular phones, notebook computers, and consumer products).

Since it is usually necessary to increase the battery or other DC input voltage of a product to a higher DC voltage or lower to a lower DC voltage, these regulators are called DC-to-DC converters. A step-down converter is used whenever the battery voltage is greater than the required load voltage. The buck converter may include an inductive switching regulator, a capacitive charge pump, and a linear regulator. On the contrary, whenever the battery voltage is lower than the required load voltage, a boost converter, usually called a boost converter, is required. The boost converter may include an inductive switching regulator or a capacitive charge pump.

However, the prior art inductive switching regulators, capacitive charge-pump converters, and linear regulators all suffer from certain limitations in terms of performance and efficiency.

<b>Inductive boost type switching converter</b>

Among the aforementioned voltage regulators, the inductive switching converter can achieve excellent performance in a wide range of current, input voltage, and output voltage. There are mainly two types of inductive switching converters-converters using single-winding inductors are usually called non-isolated converters, and converters using transformers and multi-winding inductors are usually called isolated converters. Among them, single-winding non-isolated inductive switching converters are usually used in portable products where size, efficiency, and battery life are extremely important.

Non-isolated inductive switching converters can operate in a wide range of input and output voltages and load currents with high efficiency. In particular, they are dedicated to only raising or lowering an input to a higher or lower range. Operates at low voltage. Non-isolated step-down converters are usually called decompression converters. Non-isolated boost converters are often called boost converters. The non-isolated inductive switching regulator is described in the application No. 11/890,818 known as "High-Efficiency DC/DC Voltage Converter Including Down Inductive Switching Pre-Regulator and Capacitive Switching Post-Converter" by RK Williams. The case is incorporated into this article by reference.

Inductive switching regulators suffer from various or fundamental limitations. For example, both the decompression type converter and the boost type converter exhibit the difficulty of extremely narrow pulse widths. At high conversion ratios, that is, when the output voltage is significantly different from the input voltage, narrow pulses naturally occur. Narrow pulses also occur when the output voltage is similar to the input voltage.

For example, in a boost converter, a narrow pulse width occurs whenever the required output voltage is significantly greater than the input voltage. The narrow pulse limitation makes it difficult and ineffective to raise an input voltage by a high rate-for example, by a factor of 4 or more. This occurs because a booster converter delivers an output voltage according to the following relationship in fixed frequency operation:

<maths><img file="TW201014136A_D0001.tif" /></maths>

Where V_battIs the input and D is the operating factor of the MOSFET that conducts during the magnetization period of the inductor (that is, during the magnetic energy storage period). Because of V_OUT&gt;&gt;V_batt, The expression (1-D) must be small, so D100%. As D increases, the low-side MOSFET is turned on for an increased part of the cycle, and there is less time to transfer its energy to the output capacitor. Transferring more energy in a shorter duration requires an ever-increasing current, and efficiency is compromised.

Since the low-side MOSFET must be turned off and then turned on quickly, the extremely high duty cycle in a boost converter creates a narrow turn-off pulse problem and degrades efficiency in order to achieve a high boost conversion ratio . Ideally, the converter should operate closer to a 50% duty cycle, so that a relatively equal amount of time can be used to both magnetize the inductor and then transfer the energy stored in the inductor to the output capacitor.

When the input voltage of the regulator is similar to the output voltage (that is, when the output-to-input voltage transfer ratio is close to 1), narrow pulse widths can also occur. This condition itself does not appear as a current spike, but as a phenomenon called "failure" in which the regulation is degraded.

<b>Failure in the prior art converter</b>

Regardless of whether the converter is a boost converter or a buck converter, prior art converters suffer from a problem called failure. Specifically, whenever the input voltage and output voltage are in the hundreds of millivolts (that is,<img file="TW201014136A_D0002.tif" />) When they are close to each other within the range, the adjustment ability of the converter is impaired. There can be several ways-from a single or repeated failure or discontinuity in the output voltage, a ripple increased by one of the output voltages, or a complete loss of regulation in a certain narrow voltage band to indicate the loss of regulation ability. In these situations, the converter regulation "failure".

In a booster converter that is close to failure, as D0%, whenever the duration becomes too short to achieve closed-loop control, the operating factor must be from D_minJump to 0%. With a zero operating factor, no energy is transferred from the input terminal of the converter to the inductor, so instantaneous loss of control and therefore loss of regulation. Similarly, as it switches the working factor from D_maxJump to 100%, a pressure-reducing converter loses regulation instantaneously, and when D=100%, it loses regulation completely. This is because the input terminal is basically resistively connected to the output terminal. Therefore, both the pressure-reducing configuration inductive switching regulator and the boosting configuration inductive switching regulator both suffer from failure of a conversion ratio close to 1.

Whenever the ΔV across the input and output terminals of another type of step-down converter, the linear regulator, becomes too small, the linear regulator suffers from malfunction and regulation loss. In essence, a failure occurs in a linear regulator because the loop gain of the amplifier performing the adjustment drops sharply as its transistor transmission element changes from being used as a current source to being used as a variable resistor. If the transmission element is a bipolar transistor, the V_CEGain loss occurs at a small value. In many bipolar linear regulators, this failure condition occurs below 400mV.

In the so-called "low failure" linear regulator or "LDO", a MOSFET that can operate as a current source at a lower ΔV replaces the bipolar transfer element, but because the power MOSFET transfer element is self-saturated ( That is, when the constant current operating region is transformed to its linear (ie, resistive) operating region, the linear regulator still fails at one of ΔV from 200mV to 300mV. Therefore, although the linear regulator is not switched and is not limited by the narrow pulse problem, it still suffers from a malfunction effect and a corresponding adjustment loss. In addition, the linear regulator alone can only perform pressure reduction operations.

Compare its non-isolated counterpart converters. For example, the isolated converters of flyback and forward converters can operate at a high efficiency close to 1 without switching modes or suffering failures, but their use is physically The series of large tapped inductors, coupled inductors and transformers exclude their applications in most portable products.

<b>Charge Pump Converter</b>

An alternative to a switched inductor converter is a charge pump, which uses only switches and capacitors for repeated charge redistribution (ie, continuous charging and discharging of a capacitor network driven by a clock or oscillator) A voltage conversion circuit that performs charge conversion. Although there are various prior art charge pumps using any number of flying capacitors and MOSFET switching networks, this type of converter can be pre-configured to increase or decrease a voltage but cannot be configured to Perform both step-up and step-down conversion.

The two most common topologies of boost charge pumps are charge pump voltage doubler and 1.5X fractional charge pump. For example, the charge pump voltage doubler 1 of FIG. 1A includes a voltage of V_battA battery or voltage source 2, a flying capacitor 3, MOSFETs 4, 5, 6, and 7, and an output capacitor 8. The operation of voltage doubler 1 involves in the current path<img file="TW201014136A_D0003.tif" />The flying capacitor 3 is charged successively and repeatedly in the current path, and then in the current path<img file="TW201014136A_D0004.tif" />The charge is transferred from the flying capacitor to the output capacitor 8. The charging of the flying capacitor 3 occurs by turning on the MOSFETs 4 and 5 while the MOSFETs 6 and 7 are kept off, so that after a certain time<img file="TW201014136A_D0005.tif" />, As illustrated in the equivalent circuit 10 shown in FIG. 1B. As illustrated, the voltage source 11 represents the battery 2.

The charge transfer from the capacitor 3 to the output capacitor 8 occurs by turning on the MOSFETs 6 and 7 while turning off the MOSFETs 4 and 5. The equivalent circuit 15 shown in FIG. 1C illustrates that during the charging of the output capacitor 8, the flying capacitor 3 is electrically located on the top of the battery 11 so that the voltage on the top of the battery 11 increases. due to<img file="TW201014136A_D0006.tif" />, So the capacitor 8 is charged to approximately V_battTwice the voltage, that is, the output voltage V_OUTClose to 2V_batt. Therefore, the charge pump 1 is usually called a voltage doubler. If the battery 11 has the smallest internal series resistance (not shown schematically), the charge transfer current<img file="TW201014136A_D0007.tif" />It may be sufficient to allow the voltage doubler charge pump 1 to react quickly to changing load conditions and maintain the output voltage while delivering increased current to an electrical load.

In some applications that produce an output, twice the input voltage may be too high for the electrical load being powered. In this case, the efficiency of the charge pump 1 can be very low. One way to improve the efficiency of the overall charge pump is to use a fractional charge pump 20 as shown in FIG. 2A.

As illustrated, the 1.5X charge pump 20 includes a voltage of V_battOne battery or voltage source 21, two flying capacitors 22 and 23, MOSFETs 24, 25 and 26 for charging the flying capacitors 22 and 23, and MOSFETs 27, 28, 29 and 30 for transferring charge to the output terminal , And output capacitor 31. Fractional assignments involve successive and repeated passages of current paths<img file="TW201014136A_D0008.tif" />Continuously charge the flying capacitors 22 and 23 and then go through the current path<img file="TW201014136A_D0009.tif" />The charge is transferred from the flying capacitor connected in parallel with the output capacitor 31. In particular, the charging of the flying capacitors 22 and 23 occurs by turning on the MOSFETs 24, 25, and 26 while the MOSFETs 27, 28, 29, and 30 remain off. Since the capacitors 22 and 23 are connected in series, each of the flying capacitors 22 and 23 is charged to half of the input voltage, that is,<img file="TW201014136A_D0010.tif" />. The charging conditions are illustrated by the equivalent circuit 10 shown in FIG. 2B, where the voltage source 36 represents a battery 21 that does not have any significant internal series resistance.

The transfer of charge from the capacitors 22 and 23 to the output capacitor 31 occurs by turning on the MOSFETs 27, 28, 29, and 30 while turning off the MOSFETs 24, 25, and 26. The equivalent circuit 40 shown in FIG. 2C illustrates that during the charging of the output capacitor 31, the flying capacitors 22 and 23 are connected in parallel, and the parallel combination is electrically located on the top of the battery 36 so that the voltage on the top of the battery 36 increases . due to<img file="TW201014136A_D0011.tif" />, The capacitor 31 is charged to approximately V_batt1.5 times the voltage, which is the output voltage V_OUTClose to 1.5V_batt. Therefore, the charge pump 40 is generally referred to as a fractional boost type charge pump. If the battery 36 has the smallest internal series resistance (not shown schematically), the charge transfer current<img file="TW201014136A_D0012.tif" />It may be sufficient to allow the fractional charge pump 40 to quickly respond to changing load conditions and maintain the output voltage while delivering increased current to an electrical load.

The advantage of the charge pump shown in FIGS. 1A and 2A is that under a specific voltage conversion ratio, the charge pump can exhibit a conversion efficiency close to 100%. This high efficiency occurs because very little current flows in each charge and discharge cycle.

One disadvantage of a charge pump is that it can only operate effectively at a specific conversion ratio. If the output voltage is not a multiple of the input voltage, the converter exhibits a low efficiency. If for any reason (for example, loading of output), V_outDeviation from the target voltage in the doubler by 2V_battOr a target voltage of 1.5V in a fractional charge pump_batt, The efficiency of the converter drops.

Since the efficiency of the charge pump converter decreases whenever the output-to-input voltage conversion ratio deviates from the specific voltage conversion ratios, it is impossible to generate a predetermined output voltage without significantly sacrificing efficiency.

Therefore, the charge pump only works effectively when its output voltage is a fixed fractional multiple of its input voltage. If the output voltage of a charge pump changes in proportion to its input voltage, it cannot be regarded as a voltage regulator. Adapting a charge pump to produce a fixed output voltage as the input voltage changes (for example, by partially charging a flying capacitor to force the output of the charge pump to a lower voltage) always sacrifices efficiency. Therefore, the charge pump does not become an effective voltage regulator.

<b>Limitations of prior art up-down converters</b>

In summary, the prior art DC-to-DC converters and voltage regulators suffer from several limitations as summarized in the following table.

<tables><img file="TW201014136A_D0013.tif" /></tables>

Among the available prior art converters, decompression type converters and linear regulators can only provide step-down conversion. In addition, it loses regulation whenever the input and output voltages are similar, that is, suffers from malfunction. The linear regulator also suffers from poor efficiency for large differences in input and output voltages.

Boost converters can boost an input voltage, but are subject to several limitations.

Except when<img file="TW201014136A_D0014.tif" />And because of the narrow pulse problem when the operating factor is close to 100%, it suffers from failure, and the inductive boost converter is also subject to other restrictions whenever the operating factor is close to zero. Under this condition, the ability of a boost converter to boost an input voltage by a larger multiple is limited by the extremely narrow pulses of high current, reduced efficiency, and limits its ability to adjust transients.

The charge pump that can also perform boost conversion can provide good efficiency at higher conversion ratios (for example, at 2X or 3X input) but only at precise predetermined voltage multiples. It is impractical for general voltage regulation. Any deviation from a predetermined multiple results in a significant efficiency loss.

In short, all current non-isolated converters are limited in performance whenever their output and input voltages are greatly different. Except for the charge pump that does not provide an effective adjustment method, when<img file="TW201014136A_D0015.tif" />(Ie, close to a conversion ratio of 1), the prior art DC-to-DC converter can also become unstable or lose regulation.

Since only a boost converter or a charge pump provides non-isolated boost conversion in a small space, the choice of boost conversion is even more restricted. However, boost converters suffer from high MOSFET current and low efficiency limitations at high conversion ratios. The charge pump cannot provide regulation without sacrificing efficiency.

A boost converter and voltage regulator that are effective in a wide range of input and output voltages and can achieve a high conversion ratio without operating at the limit of the duty cycle, so as to avoid the aforementioned narrow pulse problem . Ideally, such a converter should also be able to minimize the problems associated with the failure of the close-to-full voltage transfer conversion ratio. In addition, the converter should be able to supply high transient currents while maintaining tight regulation.

In a DC/DC voltage converter according to the present invention, an output terminal of a pre-regulator including an inductive switching voltage converter is connected to an input terminal of a post converter including a charge pump. The pre-regulator may include a step-down (decompression) type or a step-up (boost) type converter. The post converter can include an integral or fractional charge pump. The input terminal of the pre-regulator is the input terminal of the DC/DC voltage converter; the output terminal of the post converter is the output terminal of the DC/DC voltage converter.

According to the present invention, the post converter has a second input terminal coupled to one of the input terminals of the DC/DC converter. The second input terminal of the post converter is coupled to a terminal of a capacitor in the charge pump via a switch. In operation, the switch is repeatedly opened and closed so that the terminal of the capacitor is sequentially connected to the input terminal of the DC/DC voltage converter and disconnected from the input terminal of the DC/DC voltage converter. This is in contrast to the structure described in Application Nos. 11/890,818 and 11/890,956, in which the terminal of the capacitor in the charge pump is coupled to the output terminal of the pre-regulator via a switch.

Therefore, since the terminals of the capacitor in the charge pump are repeatedly connected to the input terminals of the DC/DC voltage converter, the voltage across the capacitor is increased to the input DC voltage instead of the intermediate voltage generated by the pre-regulator. The pre-regulator and post-converter can be driven by a common clock pulse generator, and the charge transfer from the capacitor in the charge pump to the output capacitor can occur in phase or out of phase with the magnetization of the inductor in the pre-regulator .

A DC/DC converter of the present invention can supply a relatively large transient current in response to the load demand, because the transient current performance of the converter is not affected by the series resistance of the pre-regulator.

A new series of high-efficiency DC-to-DC converters and switching regulators are disclosed in application Nos. 11/890,818, 11/890,941, 11/890,956 and 11/890,994, all of which are in The application was filed on August 8, 2007, and all of them are incorporated herein by reference. These converters exhibit dynamic up-and-down conversion performance and large voltage conversion ratio, and do not have the complicated situation of mode switching and instability under various operating conditions.

The converter disclosed therein combines an inductive energy storage element (indicated by L) and one or more continuously switched capacitive storage elements (indicated by C). In one embodiment, a two-stage voltage converter of a type of LCXU converter referred to herein includes an inductive switching pre-regulator followed by a capacitive voltage post converter. The inductive pre-regulator can increase or decrease the input voltage. The capacitive post converter boosts the voltage at its input terminal, which is the output voltage of the pre-regulator.

In a preferred embodiment, the entire two-stage converter uses synchronous switching of inductors and capacitors, and uses closed-loop feedback from the output terminal of the post converter to adjust the pulse width of the inductive pre-regulator.

In one embodiment, the so-called LCUU topology combines a boost-type inductive pre-regulator with a boost-type post-converter. The LCUU topology has a reasonable operating factor (that is, where V_in&lt;&lt;V_out) Provides a high conversion ratio for step-up voltage conversion, thereby avoiding the aforementioned narrow pulse problem of the prior art switching regulator. In a similar manner, an LCDU topology combines a step-down inductive pre-regulator with a step-up post-converter.

In the aforementioned patent disclosure, other LCXX and a related type of CLXX converter are also described. However, in particular, the present disclosure relates to a variant of the LCXU class of converters, which includes a step-up or step-down inductive pre-regulator followed by a step-up capacitive post-converter.

<b>LCXU converter job</b>

In the LCXU converter series disclosed in Application Nos. 11/890,818 and 11/890,956, the overall topology can be represented by the converter 50 shown in FIG. 3A. The converter includes a battery or power supply 51 and an inductor 53 One switch voltage pre-regulator 50A, and charge to an intermediate voltage V_yOne middle charge capacitor 54. Depending on the connection of the inductor 53, the pre-regulator 50A may include a boost converter or a buck converter.

Converter 50A output voltage V_yA voltage doubler topology is used to power a charge pump 50B. The voltage doubler topology includes a single flying capacitor 55, a network of power MOSFETs 56, 57, 58 and 59, and an output capacitor 60. The output capacitor 60 is connected in parallel with a load 61. The control circuit of the MOSFETs 56, 57, 58 and 59 (not shown) charges the capacitor 55 by turning on the MOSFETs 56 and 57 while the MOSFETs 57 and 58 are kept off, and then by turning off the MOSFETs 56 and 57 At the same time, the MOSFETs 58 and 59 are turned on to transfer the charge from the capacitor 55 to the capacitor 60.

The operating principle of the converter 50 can be achieved by charging the flying capacitor 55 to a voltage V_yIt is illustrated as the equivalent circuit 65 in FIG. 3B. In FIG. 3B, the dependent voltage source 66 represents the output voltage V of the pre-regulator 50A at the charged capacitor 54_y. During charging, a transient current<img file="TW201014136A_D0016.tif" />Flow until capacitor 55 reaches its final voltage V_y。

As shown by the equivalent circuit 70 in FIG. 3C, during the charge transfer cycle, the charge to the voltage V_yThe charged flying capacitor 55 is charged to a voltage V_yThe dependent voltage source 66 (the output voltage of the pre-regulator 50A) is on the top". Since the negative terminal of the capacitor 55 is connected to the positive terminal of the voltage source 66, the voltage increases. Then charge the capacitor 60 to a voltage of 2V_y, Is twice the output of the pre-regulator 50A. Since the output voltage of the post converter 50B is the intermediate voltage V_ySo the post converter 50B acts as a voltage doubler.

A transient current<img file="TW201014136A_D0017.tif" />It flows to transfer the charge to the capacitor 60 and to provide any current required by the electric load 61 connected in parallel with the capacitor 60. This ring affects the current<img file="TW201014136A_D0018.tif" />The series impedance includes any parasitic resistance included in the controlled voltage source 66. In other words, the transient current performance of the converter 50 is affected by the design of the pre-regulator 50A and the capacitance and type of the capacitor 54.

Another option is, as shown in the converter 80 of FIG. 4A, the post converter may include a fractional charge pump circuit. The converter 80 includes a battery or power supply 81, and an inductor 83 before switching the voltage Set the regulator 80A, and charge to an intermediate voltage V_yOne of the intermediate charging capacitors 84. Depending on the connection of the inductor 83, the pre-regulator 80A may include a boost converter or a buck converter.

Pre-regulator 80A output voltage V_yUse a fractional or 1.5X topology to power one of the charge pump 80B shown here. The fractional or 1.5X topology includes two flying capacitors 85 and 86, and power MOSFETs 87, 88, 89, 90, 91, 92, and 93 A network, and an output capacitor 94. A load 97 is connected in parallel with the output capacitor 94. The control circuit of MOSFET 87, 88, 89, 90, 91, 92, and 93 (not shown) is capacitors 85 and 86 by turning on MOSFETs 87, 88, and 89 while MOSFETs 90, 91, 92, and 93 are kept off. Charge, and switch the MOSFETs 87, 88, 89, 90, 91, 92, and 93 to transfer the charge to the capacitor 94. Charge transfer occurs by turning on MOSFETs 90, 91, 92, and 93 while biasing off MOSFETs 87, 88, and 89.

The operating principle of the fractional converter 80 can be achieved by charging each of the flying capacitors 85 and 86 to a voltage V_y/2 is illustrated as shown in the equivalent circuit 95 of FIG. 4B, where the dependent voltage source 96 represents the output of the pre-regulator 80A and the charged capacitor 84. During charging, a transient current<img file="TW201014136A_D0019.tif" />Flow until each of capacitors 85 and 86 reach equal to V_yA voltage of /2, assuming that the capacitances 85 and 86 are equal in magnitude.

As shown by the equivalent circuit 98 in FIG. 4C, during the charge transfer cycle, the charged flying capacitors 85 and 86 are connected in parallel, and their parallel combination is electrically stacked on the dependent voltage source 96 (the output voltage of the pre-regulator 80A ) On top. Since the negative terminals of capacitors 85 and 86 are connected to the positive electrons of voltage source 96, the voltage increases. Then charge the capacitor 94 to (V_y+0.5V_y) Or 1.5V_yOne voltage, which is 1.5 times the output voltage of the pre-regulator 50A. Since the output voltage is higher than the intermediate voltage V_y50% more, so post converter 80B acts as a fractional boost stage.

A transient current<img file="TW201014136A_D0020.tif" />Flow to transfer the charge to the capacitor 94 and provide any current required by an electric load 97 connected in parallel with the capacitor 94. This ring affects the current<img file="TW201014136A_D0021.tif" />The series impedance includes any parasitic resistance included in the controlled voltage source 96. In other words, the transient current performance of the circuit 80 is affected by the design of the pre-regulator 80A and the performance and type of the capacitors 85 and 86.

The inductive pre-regulators 50A and 80A in the LCXU converters 50 and 80 may include any type of DC-to-DC switching converters, but preferably include a decompression type converter or a boost type converter. In the case of a pressure reducing converter, the intermediate voltage V_yThe value is less than the input voltage V_batt, And the pre-regulator 50A or 80A reduces the supply voltage. Converters 50 and 80 are examples of the previously disclosed LCDU converters, in which the input voltage V is reduced in the first stage_batt, And the intermediate voltage V is increased in the second stage_y。

Depending on its operating conditions, this circuit can use feedback control to dynamically adapt to changing conditions to maintain an output voltage that is less than, equal to, or greater than the input voltage. In response to the feedback, fixed frequency pulse width modulation (ie, PWM or variable frequency technology) can be used to control the V of the pre-regulator_yThe output voltage.

In fixed frequency operation, the output voltage of the pressure reducing converter is given by the following expression:

<i>V</i>_<i>y</i>=<i>DV</i>_<i>batt</i>

Among them, D is the operating factor of the main switching MOSFET in the pressure reducing converter. The post converter has a transfer function of circuits 50 and 80, and the circuits 50 and 80 have a voltage transfer function given by the following expression:

<i>V</i>_<i>OUT</i>=<i>nV</i>_<i>y</i>

Where n&gt;1, that is, n=2 in the case of the voltage doubler post-converter 50B, or n=1.5 in the case of the fractional post-converter 80B. Combining these items, the entire LCDU transfer function is given by the following expression:

<i>V</i>_<i>OUT</i>=<i>nV</i>_<i>y</i>=<i>nDV</i>_<i>batt</i>

Given the value of n is 1.5 or 2, and D is in the range of 5% to 95%, the voltage conversion ratio of the converter series is V_OUT/V_battFor step-down operation, it can be less than 1, for step-up operation, it can be greater than 1, or when<img file="TW201014136A_D0022.tif" />It operates at 1 or close to 1. The LCDU converter can cover this wide range without changing the operation mode, even under the full voltage conversion condition, which suffers from instability and poor performance during the mode transition. The benefits.

Another option is that the inductive pre-regulators 50A and 80A in the LCXU converters 50 and 80 include a boost type converter. In this case, the intermediate voltage V_yThe magnitude is greater than the input voltage V_batt, And the pre-regulator 50A or 80A boosts the supply voltage. Converters 50 and 80 are examples of the previously disclosed LCUU converters, in which the input voltage V is increased in the first stage_batt, And the intermediate voltage V is increased more in the second stage_y。

Depending on its operating conditions, the LCUU circuit can use feedback control to dynamically adapt to changing conditions to maintain an output voltage that is less than, equal to, or greater than one of its inputs. In response to the feedback, fixed frequency pulse width modulation (ie, PWM or variable frequency technology) can be used to control the V of the pre-regulator_yThe output voltage.

In fixed frequency operation, the output voltage of the booster converter is given by the following expression:

<maths><img file="TW201014136A_D0023.tif" /></maths>

Among them, D is the operating factor of the main switching MOSFET rather than the synchronous rectifier MOSFET in the boost converter. As mentioned earlier, the post converter 50B or 80B has a transfer function given by the following expression:

<i>V</i>_<i>OUT</i>=<i>nV</i>_<i>y</i>

Where n&gt;1, that is, n=2 in the case of the voltage doubler post-converter 50B, or n=1.5 in the case of the fractional post-converter 80B. Combining these items, the entire LCUU transfer function is given by the following expression:

<maths><img file="TW201014136A_D0024.tif" /></maths>

Given the value of n is 1.5 or 2, and D is in the range of 5% to 95%, the voltage conversion ratio V of the LCUU converter series_OUT/V_battAlways greater than 1, which means that it can only increase the input voltage.

The advantage of the LCUU converter is that it can even achieve a large boost conversion ratio under a 50% operating factor. For example, if n=2, that is, using a voltage doubler post-converter, the voltage conversion ratio V_OUT/V_batt=4, showing the output voltage which is one-fourth of its input. In a conventional prior art booster converter, a 4X conversion ratio needs to operate at a 75% duty factor. Under a 75% operating factor, the multiplier-type LCUU converter can deliver one-eight times the output of the prior art supercharger.

One of the main advantages of operation near 50% operating factor is that the frequency of the converter can be increased without limiting the operating factor range and the size of the inductor in the pre-regulator can be reduced to avoid the narrow pulse problem described earlier. . Another advantage of close to a 50% operating factor is that the MOSFET current does not require high peak currents because there is more time available to transfer energy from the battery to the inductor and from the inductor to the output capacitor. Therefore, the LCUU converter offers several advantages compared to prior art boosted converters.

<b>Improved LCXU switching converter</b>

In the LCXU converter series, the energy transfer from the post converter to the output capacitor involves a series combination of a pre-regulator and one or more flying capacitors. For example, referring to the voltage doubler type LCXU converter 50 shown in FIG. 3A, during the charging of the output capacitor 60, the converter 50 operates as shown in the circuit 70 (FIG. 3C), where the flying capacitor 55 and the voltage The sources 66 are connected in series, and an idealized element represents the pre-regulator 50A. Current<img file="TW201014136A_D0025.tif" />It flows during the charging of the output capacitor 60 and also supplies current to any load attached to the capacitor 60 in parallel. During the alternate cycle, when the flying capacitor 55 is being charged, the output capacitor 60 must supply any current required by the load.

Ideally, current is supplied by a voltage source in series with a capacitor<img file="TW201014136A_D0026.tif" />, And so it should be able to supply high transient currents on demand without warning. However, in reality, the voltage source 66 is a decompression type converter or a boost type converter or some other DC/DC converter circuit with inherent current limitation, especially when the capacitor 54 is small. These components add series resistance to the idealized equivalent circuit 70 and limit the ability of the converter 50 to react to changes in the current demand of the load. As a result of this equivalent series parasitic resistance, transient voltage regulation can be compromised. This bad response has an adverse effect on the step load response performance of the converter 50 and can only be avoided by increasing the value of the capacitor 54 or 60.

In an LCXU converter of the present invention, the series resistance of the converter during the charge transfer period is made independent of the series resistance in the pre-regulator circuit, and the transient load current performance is improved accordingly. The new topology has a unique feature, that is, current no longer flows through the pre-regulator during discharge. Therefore, the transient voltage regulation is improved by using this technology.

An embodiment of the present invention is shown in FIG. 5A, in which a converter 100 includes a battery or power supply 101, a switching voltage converter 102 with an inductor 103, and is charged to a voltage V_yOne of the intermediate charging capacitors 104. Depending on the connection of the inductor 103, the pre-regulator 100A may include a step-up converter or a step-down converter.

Voltage V at the output of the pre-regulator 100A_yFor powering a post-charge pump 100B, in this embodiment, a network including a single flying capacitor 105, power MOSFETs 106, 107, 108, and 109, and an output capacitor 110 are used in a voltage doubler topology. The control circuit of the MOSFETs 106, 107, 108, and 109 (not shown) charges the capacitor 105 by turning on the MOSFETs 106 and 107 while the MOSFETs 108 and 109 are kept off, and then turns off the MOSFET 106 by biasing Turning on the MOSFETs 108 and 109 at the same time as and 107 transfers the charge from the capacitor 105 to the capacitor 110.

In contrast to the converter 50 shown in FIG. 3A, the converter 100 includes a MOSFET 108, one terminal of which is connected to the negative terminal of the flying capacitor 105, and a second terminal is connected to the positive terminal of the battery 101. This topology change has an obvious change in the operation of the converter 100. The change is that during the charge transfer period, the negative terminal of the flying capacitor 105 is connected to the battery voltage V via the MOSFET 108._battNot connected to the output voltage V of the DC/DC pre-regulator 100A_y。

Therefore, the converter 100 is topologically very different from the converter 50, in which the MOSFET 58 is connected to the intermediate voltage V_y. In the converter 100, instead, the corresponding MOSFET 108 is directly connected to the voltage input V_batt, Instead of connecting to the intermediate voltage V_y. The operating principle of the converter 100 can be achieved by charging the flying capacitor 105 to a voltage V_yIt is shown as an illustration in the equivalent circuit 115 of FIG. 5B, where the dependent voltage source 116 represents the output of the pre-regulator 100A and the charged capacitor 104. During charging, a transient current<img file="TW201014136A_D0027.tif" />It flows in the same way as the charging of the capacitor 54 in the converter 50 until the capacitor 105 reaches its final voltage V_y。

As shown in the equivalent circuit 118 of FIG. 5C, during the charge transfer cycle, the charged flying capacitor 105 is electrically stacked on the input voltage source V_battInstead of a non-dependent voltage source 116 (which represents the output of the pre-regulator 100A and the charged capacitor 104) on top. Since the negative terminal of the flying capacitor 105 is connected to the positive terminal of the voltage source 101, the voltage increases. Then the capacitor 110 is charged to a voltage (V_batt+V_y). This voltage is not equal to the intermediate voltage V at the output of the pre-regulator 100A_yTwice, but it is significantly larger than V_batt。

As shown in Figure 5C, the transient current<img file="TW201014136A_D0028.tif" />During the charge transfer phase, it flows to the capacitor 110 and provides any current required by an electrical load connected in parallel with the capacitor 110. The ring affects the current<img file="TW201014136A_D0029.tif" />The series impedance includes any parasitic resistance included in the battery or other voltage source 101. Therefore, the dependent voltage source 116 is not involved during the charge transfer period from the flying capacitor 105 to the output capacitor 110. The transient current performance of the converter 100 is improved because it is not dependent on the design of the pre-regulator 100A or dependent on the capacitor 104.

Therefore, according to the present invention, during the charging phase, the pre-regulator 100A is used to charge the flying capacitor 105 to an intermediate voltage V_y, And then during the charge transfer phase, the intermediate voltage V_yThe voltage V added to the battery or other voltage source 101_battTo determine the output voltage of the DC/DC converter 100. Intermediate voltage V_yThe value depends on the structure and operation of the pre-regulator 100A. However, the current does not depend on the conduction through the pre-regulator 100A during the charge transfer period.

The high transient LCXU converter 100 can be implemented as an LCDU converter, where the pre-regulator 100A is a pressure-reducing pre-regulator or a step-down pre-regulator, or another alternative system, implemented as a LCUU converter, in which the pre-regulator 100A is a boosted pre-regulator or boosted pre-regulator.

<b>Example of High Transient LCDU Converter</b>

If the disclosed LCXU converter is suitable for fixed frequency down-up operation, the pre-regulator stage exhibits a transfer characteristic V_y=DV_batt. Therefore, as shown in the equivalent circuit 120 of FIG. 6A, during the charging phase, the LCDU converter uses a dependent voltage source 121 to provide a transient current<img file="TW201014136A_D0030.tif" />Charge the flying capacitor 122 to a voltage V_y. As shown in the equivalent circuit 125 of FIG. 6B, during the charge transfer phase, the charged flying capacitor 122 is electrically stacked on the voltage V of the input voltage source 127_battOn top. Therefore, a current<img file="TW201014136A_D0031.tif" />Flow to charge the output capacitor 126 to its final value V_OUT. Since the voltage source 127 and the flying capacitor 122 are connected in series, the voltage V_OUTDepartment V_battWith V_ySum:

<i>V</i>_<i>OUT</i>=<i>V</i>_<i>batt</i>+<i>V</i>_<i>y</i>=<i>V</i>_<i>batt</i>+<i>DV</i>_<i>batt</i>=<i>V</i>_<i>batt</i>(1+<i>D</i>)

Therefore, the equivalent output-to-input voltage transfer ratio of the embodiment of the LCDU converter disclosed herein is given by the following expression:

<maths><img file="TW201014136A_D0032.tif" /></maths>

The previously disclosed 2X type LCDU converter 50 has a V_OUT/V_batt= 2D voltage transfer ratio, which means that the converter 50 can operate lower or higher than a full transfer ratio. In contrast, the LCDU version of the converter 100 always operates above a full transfer ratio. Specifically, since D varies from 0 to 100%, the transfer ratio of the LCDU version of the converter 100 varies from 1X to 2X. Therefore, although the converter involves two stages of step-down and step-up, the magnitude of the voltage doubler type post converter is greater than the step-down range of the pre-regulator and the final result is only a step-up operation.

If variable frequency control is adopted, the working factor D is determined by the quantity t_on/(t_on+t_off) Instead, where t_onIs the time period during which it is turned on to allow a magnetizing current to flow to one of the switches in the inductor 103, and t_offIt is the time period during which the switch is turned off. This allows the magnetization time of the inductor 103 to be dynamically adjusted on a cycle-by-cycle basis and the duration allowed for current recirculation (ie, when the current inductor current drops).

An embodiment of a high-transient LCDU converter 140 is shown in FIG. 6C. The converter 140 includes a low-side N-channel MOSFET 142 and is connected in series with the input voltage V_battA high-side MOSFET 141, a pre-regulator 140A, an inductor 144, and an optional capacitor 145 are connected to ground. (Note: As used in this article, the term "ground" can be different from V_battThe circuit of any voltage is grounded. ) A post converter 140B includes a flying capacitor 146, MOSFETs 147, 148, 149 and 150, and an output capacitor 151. The high-side MOSFET 141 can be connected to the gate drive circuit and the gate drive signal V_G1P-channel or N-channel with appropriate changes in polarity.

A pulse width modulation controller 152 with a clock or ramp generator 155 and a break before make (BBM) circuit 153 and 154 are used to achieve MOSFET gate driving and timing. Use the output voltage V from the converter 140_OUTThe negative feedback responds to a control voltage V_FBAchieve pulse width modulation. The level shifter 156 will V_FBThe value is adjusted to mandatory V_OUTThe appropriate voltage to become a certain target value. The PWM controller 152 may alternatively operate using variable frequency control.

The operation of the BBM circuit 153 ensures that the MOSFETs 141 and 142 are driven out of phase to avoid breakdown conduction. Specifically, the MOSFET 141 conducts to magnetize the inductor 144, that is, increases its current, and in addition, the diode 143 and the synchronous rectifier MOSFET 142 provide a current recirculation path whenever the MOSFET 141 is turned off.

Similarly, BBM circuit 154 ensures that MOSFETs 147 and 148 conduct in-phase and are driven out of phase with MOSFETs 149 and 150. In particular, both MOSFETs 147 and 148 conduct to charge the flying capacitor 146, and in addition, MOSFETs 149 and 150 conduct to transfer charge from the flying capacitor 146 to the output capacitor 151. In a preferred embodiment, both BBM circuits 153 and 154 are driven out of phase by a signal from a common clock generator 155.

In one embodiment, the inductor 144 is magnetized while charging the flying capacitor 146, thereby requiring the MOSFETs 141, 147, and 148 to be driven in phase to conduct simultaneously. In another embodiment, the inductor 144 is magnetized while transferring the charge on the flying capacitor 146 to the output capacitor 151, so that it needs to be in phase with MOSFETs 149 and 150 (that is, conduct simultaneously) and out of phase with MOSFETs 147 and 148. Drive MOSFET 141. The size of the optional capacitor 145 must be adjusted according to the gate timing and the operating current range of the converter 140.

In the monolithic implementation, the capacitor 145 partially represents the capacitor that is naturally associated with the formation of the well (that is, the PN junction used to form and integrate the MOSFETs 147 to 150).

<b>Example of High Transient LCUU Converter</b>

If the disclosed LCXU converter is suitable for a fixed frequency boost (or more accurately a boost-boost) operation, the pre-regulator stage exhibits a transfer characteristic V_y=V_batt/(1-D). Therefore, as shown in the equivalent circuit 170 of FIG. 7A, during the charging phase, the LCUU converter uses the dependent voltage source 171 to provide a transient current<img file="TW201014136A_D0033.tif" />Charge the flying capacitor 172 to a voltage V_y. During the charge transfer phase, as shown in the equivalent circuit 175 of FIG. 7B, the charged flying capacitor 172 is electrically stacked on the voltage V of the input voltage source 173_battOn top, take this a current<img file="TW201014136A_D0034.tif" />Flow to charge the output capacitor 174 to its final value V_OUT. Since the flying capacitor 172 and the voltage source 173 are connected in series, the voltage V_OUTDepartment V_battWith V_ySum:

<maths><img file="TW201014136A_D0035.tif" /></maths>

The equivalent output-to-input voltage transfer ratio of the embodiment of the LCUU converter disclosed herein is given by the following expression:

<maths><img file="TW201014136A_D0036.tif" /></maths>

The previously disclosed 2X type LCUU converter 50 has a V_OIT/V_batt=2/(1-D) voltage transfer ratio. On the contrary, the LCUU version of converter 100 is always higher than a full transfer ratio but less than the intermediate voltage V_yOperates at twice the value of one voltage. Specifically, as D changes from 0 to 75%, the transfer ratio of the LCUU version of the converter 100 changes from 2X to 6X. In the same range, the previously disclosed 2X type LCUU converter 50 will exhibit a range of 2X to 8X.

Therefore, although the converter only involves the boost phase, the transfer ratio range of the voltage doubler type post converter 50 shown in FIG. 3A is larger than that of the LCUU version of the high transient converter 100 shown in FIG. 5A Transfer ratio range. Even so, the LCUU version of converter 100 still provides an extremely wide range of transfer ratios. Working factors higher than 75% are also possible, but the current required to achieve the corresponding transfer ratio can be very high.

If variable frequency control is adopted, the working factor D is determined by the quantity t_on/(t_on+t_off) Instead, thereby allowing the magnetization time of the inductor 103 to be dynamically adjusted on a cycle-by-cycle basis and the duration for current recirculation (that is, when the current inductor current drops).

An embodiment of the high-transient LCUU converter 180 is shown in FIG. 7C. The converter 180 includes a pre-regulator 180A, which includes a low-side N-channel MOSFET 181, a floating synchronous rectifier MOSFET 183 with a corresponding inherent PN diode 184, an inductor 182 and an optional capacitor 195, and a rear The converter 180B includes a flying capacitor 185, MOSFETs 186, 187, and 188, and an output capacitor 189. The synchronous rectifier MOSFET 184 can be connected to the gate drive circuit and the gate drive signal V_G2P-channel or N-channel with appropriate changes in polarity. The MOSFET 184 serves a dual purpose in the converter 180, both as a synchronous rectifier of the pre-regulator 180A, and as one of the MOSFETs used to control time when the flying capacitor 185 is charged.

A pulse width modulation controller 190 with a clock or ramp generator 193 and a break before make (BBM) circuit 191 and 192 are used to achieve MOSFET gate driving and timing. Use the output voltage V from the converter 180_OUTThe negative feedback responds to a control voltage V_FBAchieve pulse width modulation. The level shifter 194 will V_FBThe value is adjusted to mandatory V_OUTThe appropriate voltage to become a certain target value. The PWM controller 190 may alternatively operate using variable frequency control.

The operation of the BBM circuit 191 ensures that the MOSFETs 181 and 184 are driven out of phase to avoid breakdown conduction and short-circuiting of the capacitor 195. In particular, the MOSFET 181 conducts to magnetize the inductor 182, that is, increases its current, and in addition, the diode 184 and the synchronous rectifier MOSFET 183 provide a current path for the charging capacitor 195 whenever the MOSFET 181 is turned off.

Similarly, BBM circuit 154 ensures that MOSFET 186 conducts in phase and is driven out of phase with MOSFETs 187 and 188. Specifically, the MOSFET 186 conducts to charge the flying capacitor 185 from the capacitor 195. In addition, both MOSFETs 187 and 188 are biased to conduct simultaneously to transfer charge from the flying capacitor 185 to the output capacitor 189. In a preferred embodiment, the BBM circuits 191 and 192 are driven in phase by a signal from a common clock generator 193.

In a preferred embodiment, the inductor 182 is magnetized while the flying capacitor 185 transfers its charge to the output capacitor 189, so that the MOSFETs 181, 187, and 188 need to be driven in phase to conduct simultaneously. In the opposite phase, the MOSFETs 183 and 186 are biased to conduct simultaneously, thereby charging the flying capacitor 185 to a voltage V'_y。

Since the MOSFET 183 is used both as a synchronous rectifier and for charging the flying capacitor 185, there is no stable intermediate voltage V in the converter 180 as illustrated in the circuit 100._y. Alternatively, the voltage V'_yOnly during the time when MOSFET 181 is off and MOSFETs 183 and 186 are on_y. The size of the optional capacitor 195 can be adjusted to correspond to the gate timing and the operating current range of the converter 180, but the capacitor 195 can only represent the well (that is, the PN used to form and integrate the MOSFETs 183, 186, 187, and 188). The parasitic capacitance associated with the formation of the junction).

<b>Fractional LCXU switching converter with high transient performance</b>

As mentioned earlier, in a fractional LCXU converter, the energy transfer from the converter to the output capacitor involves a series combination of a pre-regulator and one or more flying capacitors. For example, referring back to FIG. 4A, during the charging of the output capacitor 94, the fractional type LCXU converter 80 operates in one of the ways shown in the equivalent circuit 95 (FIG. 4B), in which the flying capacitors 85 and 86 and the voltage source 96 (an idealized element represents the pre-regulator 80A) in series. As shown in FIG. 4C, during the charging period of the output capacitor 94, the current<img file="TW201014136A_D0037.tif" />Flow from the parallel combination of flying capacitors 85 and 86 stacked on top of the voltage source 96 to supply current to any load connected in parallel with the output capacitor 94. During the charging phase, when the flying capacitors 85 and 86 are being charged, the output capacitor 94 must supply any current required by the load.

Ideally, current is supplied by a voltage source in series with a capacitor<img file="TW201014136A_D0038.tif" />, And so it should be able to supply high transient currents on demand without warning. However, in reality, the voltage source 96 is a decompression type converter or a boost type converter or some other DC/DC converter circuit with inherent current limitation, especially when the capacitor 84 is small. These components add series resistance to the idealized equivalent circuit 98 and limit the ability of the converter 80 to react to changes in the current needs of the load. Due to the series parasitic resistance, transient voltage regulation can be compromised. This bad response has an adverse effect on the step load response performance of the converter and can only be avoided in the LCXU converter 80 by increasing the value of the capacitor 84 or 94.

In a fractional LCXU converter of the present invention, the series resistance of the converter during the charge transfer period is made independent of the series resistance in the pre-regulator circuit, and the transient load current performance is improved accordingly. The new topology has a unique feature, that is, current no longer flows through the pre-regulator during discharge. Therefore, the transient voltage regulation is improved by using this technology.

An example of this improvement is shown in FIG. 8A, where the converter 300 includes a battery or power supply 301, a switching pre-regulator 300A with an inductor 303, and is charged to a voltage V_yOne of the intermediate charging capacitors 304. Depending on the connection of the inductor 303, the pre-regulator 300A may include a boost converter or a buck converter.

Converter 300A output voltage V_yCharge a part of a post converter 300B, which includes a charge pump using a fractional topology, the fractional topology includes: two flying capacitors 305 and 306, a network of power MOSFETs 307 to 313, and an output Capacitor 314. The control circuit of the MOSFETs 307 to 313 (not shown) charges the flying capacitors 305 and 306 by turning on the MOSFETs 307, 308, and 309 while the MOSFETs 310, 311, 312, and 313 remain off, and then charges the flying capacitors 305 and 306 by biasing The MOSFETs 307, 308, and 309 are turned off while the MOSFETs 310, 311, 312, and 313 are turned on to transfer the charge on the flying capacitors 305 and 306 to the capacitor 314.

In contrast to the converter 80 shown in FIG. 4A, in the converter 300, the negative terminals of the flying capacitors 305 and 306 are connected to the positive terminal of the battery 301 via MOSFETs 310 and 311, respectively. This topology change forms one of the significant changes in the operation of the converter 300. During the charge transfer period, MOSFET 310 and 311 are connected to the battery voltage V_battInstead of connecting to the output V of the pre-regulator 300A_y。

The circuit 300 is completely different from the converter 80 in topology, in which the MOSFETs 90 and 91 are connected to the voltage V_y. In converter 300, MOSFETs 90 and 91 are replaced by MOSFETs 310 and 311, and MOSFETs 310 and 311 are connected to the input voltage V_batt, Not the intermediate voltage V_y. The operating principle of the converter 300 can be achieved by charging the flying capacitors 305 and 306 connected in series to a voltage V_yIt is illustrated as shown in the equivalent circuit 320 of FIG. 8B, where the dependent voltage source 321 represents the output of the pre-regulator 300A and the charged capacitor 304. During the charging period of the flying capacitors 305 and 306, a transient current<img file="TW201014136A_D0039.tif" />Flow until each of the capacitors 305 and 306 reach the voltage V in the same way as the charging of the flying capacitors 85 and 86 in the converter 80_y/2。

As shown in the equivalent circuit 325 of FIG. 8C, during the charge transfer cycle, the charged flying capacitors 305 and 306 are stacked on the input voltage source V_battThe parallel combination on the top of the non-dependent voltage source 321 is electrically connected in parallel. Since the negative terminals of the flying capacitors 305 and 306 are connected to the positive terminal of the voltage source 301, the voltage increases. Then charge the capacitor 314 to a voltage (V_batt+V_y/2). The voltage is not equal to 1.5 times the output of the pre-regulator 300A, but it is significantly greater than V_batt。

Transient current<img file="TW201014136A_D0040.tif" />During the charge transfer phase, it flows to the capacitor 314 to provide any current required by an electrical load connected in parallel with the capacitor 314. The ring affects the current<img file="TW201014136A_D0041.tif" />The series impedance includes any parasitic resistance contained in the battery or the independent voltage source 301. The dependent voltage source 321 is not involved during the charge transfer period from the flying capacitors 305 and 306 to the output capacitor 314. The transient current performance of the circuit 326 is improved because it is not dependent on the design of the pre-regulator 300A or dependent on the capacitor 304.

Therefore, according to the present invention, during the charging phase, the pre-regulator 300A is used to charge each of the flying capacitors to a voltage V_y/2, and then during the charge transfer phase, the voltage V_y/2 added to the voltage V of the battery or other voltage source 301_battTo determine the output voltage of the DC/DC converter 300. Voltage V_yThe value depends on the structure and operation of the pre-regulator 300A. However, the current does not depend on the conduction through the pre-regulator 300A during the charge transfer period.

The high transient fraction LCXU converter 300 can be implemented as an LCDU converter, where the pre-regulator 300A is a pressure-reducing pre-regulator or a step-down pre-regulator, or another alternative system, implemented as An LCUU converter, in which the pre-regulator 300A is a supercharged pre-regulator or a boost pre-regulator.

<b>Example of LCDU converter capable of high transient fraction</b>

If the disclosed LCXU converter is suitable for fixed frequency down-up operation, the pre-regulator stage exhibits a transfer characteristic V_y=DV_batt. Therefore, as shown in the equivalent circuit 350 of FIG. 9A, during the charging phase, the LCDU converter uses a dependent voltage source 351 for transient current<img file="TW201014136A_D0042.tif" />Charge each of the flying capacitors 352 and 353 to a voltage V_y/2. As shown in the equivalent circuit 355 of FIG. 9B, during the charge transfer phase 355, the charged flying capacitors 352 and 353 are connected in parallel, and the parallel combination is electrically stacked on the voltage V of the input voltage source 356_battOn top. Therefore, a current<img file="TW201014136A_D0043.tif" />Flow to charge the output capacitor 357 to its final value V_OUT. Since the parallel combination of the voltage source 356 and the flying capacitors 352 and 353 is connected in series, the voltage V_OUTDepartment V_battWith V_ySum of /2:

<maths><img file="TW201014136A_D0044.tif" /></maths>

Therefore, the equivalent output-to-input voltage transfer ratio of the embodiment of the fractional LCDU converter disclosed herein is given by the following expression:

<maths><img file="TW201014136A_D0045.tif" /></maths>

The previously disclosed 1.5X type LCDU converter 80 has V_OUT/V_batt= A voltage transfer ratio of 1.5D, which means that the converter 80 can operate at a full transfer ratio lower or higher. In contrast, the LCDU version of converter 300 always operates above a full transfer ratio. Specifically, since D varies from 0 to 100%, the shift ratio of the LCDU version of the converter 300 varies from 1X to 1.5X. Therefore, although the converter involves both the step-down stage and the step-up stage, the magnitude of the 1.5X type post-converter is greater than the step-down range of the pre-regulator and the final result is only the step-up operation.

If variable frequency control is adopted, the working factor D is determined by the quantity t_on/(t_on+t_off) Instead, thereby allowing the magnetization time of the inductor 303 to be dynamically adjusted on a cycle-by-cycle basis and the duration for current recirculation (that is, when the current inductor current drops).

An embodiment of a high transient fraction LCDU converter 370 is shown in FIG. 9C. The converter 370 includes a pre-regulator 370A including a low-side N-channel MOSFET 372, a high-side MOSFET 371, an inductor 374, and an optional capacitor 375. A post converter 370B includes flying capacitors 376 and 377, MOSFETs 378, 379, 380, 381, 382, 383, and 384, and an output capacitor 385. The high-side MOSFET 371 can be connected to the gate drive circuit and the gate drive signal V_G2P-channel or N-channel with appropriate changes in polarity.

A pulse width modulation controller 386 with a clock or ramp generator 389 and break before make (BBM) circuits 387 and 388 are used to achieve MOSFET gate driving and timing. Use the output voltage V from the converter 370_OUTThe negative feedback responds to a control voltage V_FBAchieve pulse width modulation. Level shifter 390 will V_FBThe value is adjusted to mandatory V_OUTThe appropriate voltage to become a certain target value. The PWM controller 386 may alternatively operate using variable frequency control.

The operation of the BBM circuit 387 ensures that the MOSFETs 371 and 372 are driven out of phase to avoid breakdown conduction. Specifically, the MOSFET 371 conducts to magnetize the inductor 374, that is, increases its current, and in addition, the diode 373 and the synchronous rectifier 372 provide a current recirculation path whenever the MOSFET 371 is turned off.

Similarly, BBM circuit 388 ensures that MOSFETs 378, 379, and 380 conduct in-phase and are driven out of phase with MOSFETs 381, 382, 383, and 384. In particular, the MOSFETs 378, 379, and 380 conduct at the same time to charge the flying capacitors 376 and 377, and in addition, the MOSFETs 381, 382, 383, and 384 conduct to transfer the charge from the flying capacitors 376 and 377 to the output capacitor 385. In a preferred embodiment, BBM circuits 387 and 388 are driven in phase by a signal from a common clock generator 389.

In one embodiment, the inductor 374 is magnetized while charging the flying capacitors 376 and 377, so that the MOSFETs 371, 378, 379, and 380 need to be driven in phase to conduct simultaneously. In another embodiment, the inductor 374 is magnetized while transferring the charge on the flying capacitors 376 and 377 to the output capacitor 385, so that it needs to be in phase with the MOSFETs 381 to 384 (that is, conduct simultaneously) and be in phase with the MOSFETs 378 to 380. The MOSFET 371 is driven out of phase. The size of the optional capacitor 375 must be adjusted to correspond to the gate timing and the operating current range of the converter 370.

In the monolithic implementation, the capacitor 375 partially represents the capacitor that is naturally associated with the formation of the well (ie, the PN junction used to form and integrate the MOSFETs 378 to 384).

<b>Example of LCUU converter capable of high transient score</b>

If the disclosed LCXU converter is suitable for a fixed frequency boost (or more accurately a boost-boost operation), the pre-regulator stage exhibits a transfer characteristic V_y=V_batt/(1-D). Therefore, as shown in the equivalent circuit 400 of FIG. 10A, during the charging phase, the LCUU converter uses a dependent voltage source 401 to provide a transient current<img file="TW201014136A_D0046.tif" />Charge each of the flying capacitors 402 and 403 to a voltage V_y/2 or V_batt/2(1-D). As shown in the equivalent circuit 405 of FIG. 10B, during the charge transfer phase, the charged flying capacitors 402 and 403 are connected in parallel, and the parallel combination is electrically stacked on the voltage V of the input voltage source 406_battOn top, take this a current<img file="TW201014136A_D0047.tif" />Flow to charge the output capacitor 407 to its final value V_OUT. Since the parallel combination of the flying capacitors 402 and 403 is connected in series with the voltage source 406, the voltage V_OUTDepartment V_battWith V_ySum:

<maths><img file="TW201014136A_D0048.tif" /></maths>

The equivalent output-to-input voltage transfer ratio of the embodiment of the LCUU converter disclosed herein is given by the following expression:

<maths><img file="TW201014136A_D0049.tif" /></maths>

The previously disclosed 1.5X type LCUU converter 80 has a V_OUT/V_batt=1.5/(1-D) voltage transfer ratio. On the contrary, the LCUU version of converter 300 is always higher than 1 conversion but less than the voltage V_yIt operates at 1.5 times the voltage. Specifically, as D changes from 0 to 75%, the transfer ratio of the LCUU version of the converter 300 changes from 1.5X to 4X. In the same range, the previously disclosed 1.5X type LCUU converter 80 will exhibit a range of 1.5X to 6X.

Therefore, although the converter involves two boost-only stages, the transfer ratio range of the fractional post-converter 80 shown in FIG. 4A is larger than that of the LCUU version of the high transient converter 300 shown in FIG. 8A Boost range. However, the LCUU version of converter 300 still provides an extremely wide range of transfer ratios. Working factors higher than 75% are also possible, but the current required to achieve the corresponding transfer ratio can be quite high.

If variable frequency control is adopted, the working factor D is determined by the quantity t_on/(t_on+t_off) Instead, thereby allowing the magnetization time of the inductor 303 to be dynamically adjusted on a cycle-by-cycle basis and the duration for current recirculation (that is, when the current inductor current drops).

An embodiment of the high transient score LCUU converter 420 is shown in FIG. 10C. The converter 420 includes a pre-regulator 420A, which includes a low-side N-channel MOSFET 421, a floating synchronous rectifier MOSFET 423 with a corresponding inherent PN diode 424, an inductor 422 and an optional capacitor 445, and a rear The converter 420B includes flying capacitors 425 and 426, MOSFETs 425 to 432, and an output capacitor 433. The synchronous rectifier MOSFET 423 can be connected to the gate drive circuit and the gate drive signal V_G2P-channel or N-channel with appropriate changes in polarity. The MOSFET 423 serves a dual purpose in the converter 420, both as a synchronous rectifier of the pre-regulator 420A, and as one of the MOSFETs used to control time when charging the flying capacitors 425 and 426.

A pulse width modulation controller 434 with a clock or ramp generator 437 and a break before make (BBM) circuit 435 and 436 are used to achieve MOSFET gate driving and timing. Use the output voltage V from the converter 420_OUTThe negative feedback responds to a control voltage V_FBAchieve pulse width modulation. The level shifter 438 divides V_FBThe value is adjusted to mandatory V_OUTThe appropriate voltage to become a certain target value. The PWM controller 434 may alternatively operate using variable frequency control.

The operation of the BBM circuit 435 ensures that the MOSFETs 421 and 423 are driven out of phase to avoid breakdown conduction and short-circuiting of the capacitors 425 and 426. In particular, the MOSFET 421 conducts to magnetize the inductor 422, that is, increases its current. In addition, the diode 424 and the synchronous rectifier MOSFET 423 provide a current path for charging the capacitors 425 and 426 whenever the MOSFET 421 is turned off. .

Similarly, the BBM circuit 436 ensures that the MOSFET 423 conducts in-phase with the MOSFETs 427 and 428 and is driven out of phase with the MOSFETs 429, 430, 431, and 432. In particular, the MOSFETs 423, 427, and 428 conduct to charge the flying capacitors 425 and 426 from the capacitor 445. In addition, the MOSFETs 429 to 432 are biased to conduct simultaneously to transfer charge from the flying capacitors 425 and 426 to the output capacitor 453. In a preferred embodiment, the BBM circuits 435 and 436 are driven in phase by a signal from a common clock generator 437.

In a preferred embodiment, the inductor 422 is magnetized while the flying capacitors 425 and 426 transfer their charge to the output capacitor 453, so that the MOSFETs 429, 430, 431, and 432 need to be driven in phase to conduct simultaneously. In the reverse phase, the MOSFETs 423, 427, and 428 are biased to conduct simultaneously, thereby charging each of the flying capacitors 425 and 426 to a voltage V'_y/2。

Since the MOSFET 423 is used both as a synchronous rectifier and for charging the flying capacitors 425 and 426, as illustrated in the circuit 300, there is no stable intermediate voltage V in the converter 420_y. Alternatively, the voltage V'_yIt appears as V only during the time when MOSFET 421 is off and MOSFETs 423, 427 and 428 are on_y. The size of the optional capacitor 445 can be adjusted to correspond to the gate timing and the operating current range of the converter 420, but the capacitor 445 can only be used to form and integrate the PN junction of MOSFETs 424 to 433. Form the associated parasitic capacitance.

The embodiments described above are illustrative and not restrictive. Based on the above description, those skilled in the art will understand many additional and alternative embodiments.

<p>1. . . Charge Pump Voltage Doubler</p><p>2. . . Battery or voltage source</p><p>3. . . Flying capacitor</p><p>4. . . MOSFET</p><p>5. . . MOSFET</p><p>6. . . MOSFET</p><p>7. . . MOSFET</p><p>8. . . Output capacitor</p><p>10. . . Equivalent Circuit</p><p>11. . . power source</p><p>15. . . Equivalent Circuit</p><p>20. . . Fractional Charge Pump</p><p>twenty one. . . Battery or voltage source</p><p>twenty two. . . Flying capacitor</p><p>twenty three. . . Flying capacitor</p><p>twenty four. . . MOSFET</p><p>25. . . MOSFET</p><p>26. . . MOSFET</p><p>27. . . MOSFET</p><p>28. . . MOSFET</p><p>29. . . MOSFET</p><p>30. . . MOSFET</p><p>31. . . Output capacitor</p><p>36. . . power source</p><p>40. . . Equivalent Circuit</p><p>50. . . converter</p><p>51. . . Battery or power supply</p><p>52. . . Pulse width modulation controller</p><p>53. . . Break before Make (BBM) circuit</p><p>54. . . Break before Make (BBM) circuit</p><p>55. . . Single flying capacitor</p><p>56. . . Power MOSFET</p><p>57. . . Power MOSFET</p><p>58. . . Power MOSFET</p><p>59. . . Power MOSFET</p><p>60. . . Output capacitor</p><p>65. . . Equivalent Circuit</p><p>66. . . Dependent voltage source</p><p>70. . . Equivalent Circuit</p><p>80. . . converter</p><p>81. . . Battery or power supply</p><p>83. . . Inductor</p><p>84. . . Intermediate charging capacitor</p><p>85. . . Flying capacitor</p><p>86. . . Flying capacitor</p><p>87. . . Power MOSFET</p><p>88. . . Power MOSFET</p><p>89. . . Power MOSFET</p><p>90. . . Power MOSFET</p><p>91. . . Power MOSFET</p><p>92. . . Power MOSFET</p><p>93. . . Power MOSFET</p><p>94. . . Output capacitor</p><p>95. . . Equivalent Circuit</p><p>96. . . Dependent voltage source</p><p>98. . . Equivalent Circuit</p><p>100. . . converter</p><p>101. . . Battery or power supply</p><p>102. . . Switching voltage converter</p><p>103. . . Inductor</p><p>104. . . Intermediate charging capacitor</p><p>105. . . Single flying capacitor</p><p>106. . . Power MOSFET</p><p>107. . . Power MOSFET</p><p>108. . . Power MOSFET</p><p>109. . . Power MOSFET</p><p>110. . . Output capacitor</p><p>115. . . Equivalent Circuit</p><p>116. . . Dependent voltage source</p><p>118. . . Equivalent Circuit</p><p>120. . . Equivalent Circuit</p><p>121. . . Dependent voltage source</p><p>122. . . Flying capacitor</p><p>125. . . Equivalent Circuit</p><p>126. . . Output capacitor</p><p>127. . . Input voltage source</p><p>140. . . High transient LCDU converter</p><p>141. . . High-side MOSFET</p><p>142. . . Low-side N-channel MOSFET</p><p>143. . . Diode</p><p>144. . . Inductor</p><p>145. . . Optional capacitor</p><p>146. . . Flying capacitor</p><p>147. . . MOSFET</p><p>148. . . MOSFET</p><p>149. . . MOSFET</p><p>150. . . MOSFET</p><p>151. . . Output capacitor</p><p>152. . . Pulse width modulation controller</p><p>153. . . Break before Make (BBM) circuit</p><p>154. . . Break before Make (BBM) circuit</p><p>155. . . Has a clock or ramp generator</p><p>156. . . Level shifter</p><p>171. . . Dependent voltage source</p><p>172. . . Flying capacitor</p><p>173. . . Input voltage source</p><p>174. . . Output capacitor</p><p>175. . . Equivalent Circuit</p><p>180. . . High transient LCUU converter</p><p>181. . . Low-side N-channel MOSFET</p><p>182. . . Inductor</p><p>183. . . Floating synchronous rectifier MOSFET</p><p>184. . . PN diode</p><p>185. . . Flying capacitor</p><p>186. . . MOSFET</p><p>187. . . MOSFET</p><p>188. . . MOSFET</p><p>189. . . Output capacitor</p><p>190. . . Pulse width modulation controller</p><p>191. . . Break before Make (BBM) circuit</p><p>192. . . Break before Make (BBM) circuit</p><p>193. . . Clock or ramp generator</p><p>194. . . Level shifter</p><p>300. . . converter</p><p>301. . . Battery or power supply</p><p>303. . . Inductor</p><p>304. . . Intermediate charging capacitor</p><p>305. . . Flying capacitor</p><p>306. . . Flying capacitor</p><p>307. . . Power MOSFET</p><p>308. . . Power MOSFET</p><p>309. . . Power MOSFET</p><p>310. . . Power MOSFET</p><p>311. . . Power MOSFET</p><p>312. . . Power MOSFET</p><p>313. . . Power MOSFET</p><p>314. . . Output capacitor</p><p>320. . . Equivalent Circuit</p><p>321. . . Dependent voltage source</p><p>325. . . Equivalent Circuit</p><p>326. . . Circuit</p><p>350. . . Equivalent Circuit</p><p>351. . . Dependent voltage source</p><p>352. . . Flying capacitor</p><p>353. . . Flying capacitor</p><p>355. . . Charge transfer stage</p><p>356. . . Input voltage source</p><p>357. . . Output capacitor</p><p>370. . . High transient fraction LCDU converter</p><p>371. . . High-side MOSFET</p><p>372. . . Low-side N-channel MOSFET</p><p>373. . . Diode</p><p>374. . . Inductor</p><p>375. . . Optional capacitor</p><p>376. . . Flying capacitor</p><p>377. . . Flying capacitor</p><p>378. . . MOSFET</p><p>379. . . MOSFET</p><p>380. . . MOSFET</p><p>381. . . MOSFET</p><p>382. . . MOSFET</p><p>383. . . MOSFET</p><p>384. . . MOSFET</p><p>385. . . Output capacitor</p><p>386. . . Pulse width modulation controller</p><p>387. . . Break before Make (BBM) circuit</p><p>388. . . Break before Make (BBM) circuit</p><p>389. . . Clock or ramp generator</p><p>390. . . Level shifter</p><p>400. . . Equivalent Circuit</p><p>401. . . Dependent voltage source</p><p>402. . . Flying capacitor</p><p>403. . . Flying capacitor</p><p>405. . . Equivalent Circuit</p><p>406. . . Input voltage source</p><p>407. . . Output capacitor</p><p>420. . . High transient score LCUU converter</p><p>421. . . Low-side N-channel MOSFET</p><p>422. . . Inductor</p><p>423. . . Floating synchronous rectifier MOSFET</p><p>424. . . PN diode</p><p>425. . . Flying capacitor (MOSFET)</p><p>426. . . Flying capacitor MOSFET</p><p>427. . . MOSFET</p><p>428. . . MOSFET</p><p>429. . . MOSFET</p><p>430. . . MOSFET</p><p>431. . . MOSFET</p><p>432. . . MOSFET</p><p>433. . . Output capacitor</p><p>434. . . Pulse width modulation controller</p><p>435. . . Break before Make (BBM) circuit</p><p>436. . . Break before Make (BBM) circuit</p><p>437. . . Clock or ramp generator</p><p>438. . . Level shifter</p><p><img file="TW201014136A_D0050.tif" />. . . Transient current/current path</p><p><img file="TW201014136A_D0051.tif" />. . . Current path/transient current/charge transfer current</p>

Figure 1A is a circuit diagram of a charge pump voltage doubler;

Figures 1B and 1C are respectively equivalent circuit diagrams of the charge pump voltage doubler during the charging and charge transfer phases;

Figure 2A is a circuit diagram of a fractional 1.5X charge pump;

Figures 2B and 2C are respectively equivalent circuit diagrams of the fractional 1.5X charge pump during the charging and charge transfer phases;

Figure 3A is a circuit diagram of an LCXU converter with a 2X post converter;

3B and 3C are respectively equivalent circuit diagrams of the LCXU converter of FIG. 3A during the charging and charge transfer phases;

Figure 4A is a circuit diagram of an LCXU converter with a 1.5X post converter;

4B and 4C are respectively equivalent circuit diagrams of the LCXU converter of FIG. 4A during the charging and charge transfer phases;

Figure 5A is a circuit diagram of an LCXU converter with a 2X post converter according to the present invention;

5B and 5C are respectively equivalent circuit diagrams of the LCXU converter of FIG. 5A during the charging and charge transfer phases;

6A and 6B are respectively equivalent circuit diagrams of an LCDU converter with a 2X post converter during the charging and charge transfer phases;

Fig. 6C is a circuit diagram of the LCDU converter with 2X post converter;

7A and 7B are respectively equivalent circuit diagrams of an LCUU converter with a 2X post converter during the charging and charge transfer phases;

Figure 7C is a circuit diagram of the LCUU converter with a 2X post converter;

8A is a circuit diagram of an LCXU converter with a 1.5X post converter according to the present invention;

8B and 8C are respectively equivalent circuit diagrams of the LCXU converter of FIG. 8A during the charging and charge transfer phases;

9A and 9B are respectively equivalent circuit diagrams of an LCDU converter with a 1.5X post converter during the charging and charge transfer phases;

Figure 9C is a circuit diagram of the LCDU converter with a 1.5X post converter;

10A and 10B are respectively equivalent circuit diagrams of an LCUU converter with a 1.5X post converter during the charging and charge transfer phases; and

Figure 10C is a circuit diagram of the LCUU converter with a 1.5X post converter.