Signal Output Apparatus, a Charge Pump, a Voltage Doubler and a Method to Output Current

FIELD OF THE INVENTION

The present invention relates to the field of electronic circuit technologies, and in particular, to a signal output apparatus, a charge pump, a voltage doubler and a method for outputting a current.

BACKGROUND OF THE INVENTION

A charge pump circuit is a circuit for generating a voltage higher than its input voltage. In some interface circuits, an output voltage higher than the input voltage and a large output current are required to be provided, thus a charge pump-type booster circuit, which can generate a large current output, is needed in the interface circuit chip. For example, in a Universal Serial Bus On-the-Go (USB OTG) interface circuit, a vbus in a cable is required to be driven to a voltage over 4.75V, and a current over 8 mA should be provided to a Self-Powered Device or a current higher than 100 mA should be provided to a Bus-Powered Device.

In A High-Efficiency CMOS Voltage Doubler, Pierre Favrat, et al., IEEE JOURNAL OF SOLID-STATE CIRCUITS, Vol. 33, No. 3, MARCH 1998, a structure of a charge pump is described. FIG. 1 is a schematic diagram of the circuit structure of the charge pump, in which a clock signal CK changes periodically between an input voltage V_INand 0 and a CKN is the inverted clock of CK. After an operation time long enough, an output voltage V_OUTwill be finally stabilized at about 2×V_IN. However, when a current with order of magnitude of tens or even hundreds mA is provided by using the charge pump shown in FIG. 1, the capacitance of the capacitor in the charge pump will be very large (with the order of magnitude of about 0.1 g), thereby causing that the capacitor only can be connected to the charge pump circuit in a chip pin mode from the outside of the charge pump circuit chip, thus one or more of nodes CP1 and CP2 will be pins for the chip to connect to the outside capacitor. In the practical application using standard CMOS manufacturing process, N-type metal oxide semiconductor (NMOS) N1a and N2a may be destroyed due to electrostatic discharge (ESD) or latchup, because their bulks have no independent wells for protection, and their sources or drains and their gates are connected to high voltage chip pins.

In A Regulated Charge Pump With Small Ripple Voltage and Fast Start-Up, Jae-Youl Lee, et al., IEEE JOURNAL OF SOLID-STATE CIRCUITS, Vol. 41, NO. 2, FEBRUARY 2006, another structure of a charge pump is disclosed. FIG. 2 is a schematic diagram showing the circuit structure of the charge pump. There exist the following problems in the structure of this charge pump.

1) The drain and gate of a P-type metal oxide semiconductor (PMOS) P2b are connected with each other, which results in that, when a capacitor C2b is charged, the ascending extent of voltage on a node CP2 is limited, and the charges that can be obtained by the charge pump from the input voltage V_INin each period is limited. Therefore, such a structure greatly limits the quantity of charges that can be output from the node CP2 at a high potential (about 2×V_IN). If providing a large current output by using such a structure, a chip with a large area is needed, and the cost is relatively high.

2) Because a bulk of an output switch P3b is connected with an output node V_OUT, when other circuits connected with the V_OUTrequires to drive V_OUTto logic 0 level when the charge pump is not at work, the PN junction between the P-type source (connected with the node CP2) and the N well (connected with the bulk) of the P3b will be forward conducting, thus the potential of V_OUTcan only be driven to V_IN-Vth_PMOS-Vth_PNat the lowest (where Vth_PMOSis the cut-in voltage of the P-type metal oxide semiconductor, and Vth_PNis the forward conducting voltage of the PN junction), and cannot be normally driven to logic 0 level by the external circuit.

CN Patent NO. ZL 02157151.1 discloses a charge pump and a voltage doubler using the same. A schematic diagram of the whole circuit structure of the voltage doubler is as shown in FIG. 3, where CK and CKN are a pair of oscillation signals with opposite phases and a swing of 0˜V_IN. The operating principle of the charge pump circuit is as follows.

It is hypothesized that, at the initial state, the potential of the node CKN is V_INand the potential of the node CK is 0, then a module 2 outputs a hclk1 (the initial potential of a signal hclk2 output by the module 2 is about V_IN) signal with the potential of 0 to a module 1, so that the potential of the output signal CP1 of module 1 is pulled up to V_INvia a switch P1c, while the potential of the output signal CP2 of module 1 is raised to V_INdue to the capacitance characteristic. Then the potential of the node CKN is inverted to 0 and the potential of the node CK is inverted to V_IN, and the potential of the output signal CP1 of module 1 is raised to 2×V_INdue to the capacitance characteristic, the module 1 outputs the CP1 signal with the potential of 2×V_INto the module 2, the module 2 generates an hclk1 signal with a potential of 2×V_INby using the CP1 signal with the potential of 2×V_INoutput by the module 1 and outputs the hclk1 signal to the module 1, so that the CP1 signal output by the module 1 keeps at the potential of 2×V_IN; furthermore, the module 2 further outputs an hclk2 signal with a potential of 0 to the module 1, so that the CP2 signal output by the module 1 is again raised to the potential of V_IN(at the moment the potential of the node CKN is inverted to 0, the potential of the CP2 signal output by the module 1 is pulled down to 0 due to the capacitance characteristic). Subsequently, by the same token, when the potentials of the nodes CK and CKN are inverted between 0 and V_IN, two output signals CP1 and CP2 with opposite phases may vary periodically in a range of V_INto 2×V_IN.

The solution described in CN Patent NO. ZL 02157151.1 has the following disadvantages:

1) In addition to providing charges to the V_OUT, the output signals CP1 and CP2 of the module 1 act as the input signals of the module 2 to make the potentials of the output signals hclk1 and hclk2 of the module 2 (at the same time, they are also the input signals of the module 1) change from 0 to 2×V_INand to make the potentials of the nodes a and b of the module 2 change from 0 to V_IN, thus the charges stored by the capacitors C1c and C2c are wasted and the output charges are reduced;

2) When a circuit is required to output a large current (a current of tens or even hundreds mA) when the output voltage V_OUTis higher than the input voltage V_IN, the capacitance of the capacitor C1c or C2c would be of the order of magnitude of about 0.1 uF, thus the capacitor can only be connected from the outside of the circuit chip. At this point, the upper and lower plates of the capacitor and the V_OUTare high voltage input/output pins of the chip. The metal oxide semiconductor would readily subject to the destruction of electrostatic discharge or latchup when the high voltage input/output pins (for example, CP1, CP2 and V_OUT) of the chip are connected directly to the gate, the source and the drain of the metal oxide semiconductor (MOS) simultaneously;

3) Because the CK and CKN are an operation clock signal and an inverted operation clock signal with an opposite phase to the operation clock signal, respectively, the charge loss due to the clock level inversion cannot be avoided, and the efficiency of the energy output of the V_OUTis relatively low. Especially, when the clock signal is inverted, a problem of charge loss may arise because a breakover current from the V_INto the GND may flow through a driver that provides charges to the lower plate of the capacitor C1c or C2c; and when the clock signal is inverted, a problem of charge loss may arise because the switches P6c and P5c and the switches P1c and P8c are conducting at the same time, which causes the charges to flow back from the CP1 (or CP2) with a high potential to the CP2 (or CP1) with a low potential. Furthermore, because the current that flows through each metal oxide semiconductor (MOS) is large in the application of outputting a large current, and the parasitic capacitance is relatively large due to the size of each MOS is relatively large, thus the ascending time and the descending time of the clock signal level inversion is relatively long, thereby causing the above two problems of charge loss to be serious. Moreover, because it is limited by the structural characteristic of this solution, even if the CK and CKN are modified as two-phase non-overlap clock signals, the above two problems of charge loss still cannot be solved at the same time.

SUMMARY OF THE INVENTION

The embodiments of the invention provide a signal output apparatus, a charge pump, a voltage doubler and a method for outputting a current, thereby avoid that, when a large current is output in the case that the output voltage V_OUTis higher than the input voltage V_IN, the MOS is fragile to be destroyed by the electrostatic discharge or latchup and the efficiency of energy output of the circuit is low in the prior art.

An embodiment of the invention provides a signal output apparatus, including:

a first signal generating unit, adapted to generate a first oscillation signal with a swing of input voltage V_INto 2×V_INand a second oscillation signal with a swing of 0 to V_INand a phase the same as that of the first oscillation signal; and

a second signal generating unit, adapted to take the first oscillation signal and the second oscillation signal as triggering signals, convert the input voltage to a first output signal with a swing of 0 to 2×V_INand output the first output signal, wherein the phases of the first output signal and the first oscillation signal are the same.

An embodiment of the invention provides a charge pump, including:

a signal output module, adapted to output a first output signal with a swing of 0 to 2×V_IN; and

a charge storage module, adapted to take the first output signal as a triggering signal to generate two states, wherein one state is to charge one plate of a capacitor by an input voltage V_IN, and the other state is to provide a voltage higher than 0 volt to the other plate of the capacitor.

An embodiment of the invention provides a voltage doubler, including:

a signal output module, adapted to output a first output signal with a swing of 0 to 2×V_IN, and a first oscillation signal and a third oscillation signal with the same swing of V_INto 2×V_INbut with opposite phases;

a charge storage module, adapted to take the first output signal as a triggering signal to generate two states, wherein one state is to charge one plate of a first capacitor by an input voltage V_IN, and the other state is to provide a voltage higher than 0 volt to the other plate of the first capacitor; and

a current output switch module, adapted to take the first oscillation signal and the third oscillation signal as triggering signals and output the charges stored in the charge storage module via the first capacitor.

An embodiment of the invention provides a method for outputting a current, including:

generating a first oscillation signal with a swing of V_INto 2×V_IN, a second oscillation signal with a swing of 0 to V_INand a phase the same as that of the first oscillation signal, and a third oscillation signal with a swing the same as that of the first oscillation signal but with an opposite phase;

taking the first oscillation signal and the second oscillation signal as triggering signals, converting the input voltage to a first output signal with a swing of 0 to 2×V_INand outputting the first output signal, wherein the phases of the first output signal and the first oscillation signal are the same;

taking the first output signal as a triggering signal to generate two states, wherein one state is to charge one plate of a capacitor by the input voltage V_IN, and the other state is to provide a voltage higher than 0 volt to the other plate of the capacitor; and

taking the first oscillation signal and the third oscillation signal as triggering signals to output the charges on the plate of the capacitor.

In the solutions provided in the embodiments of the invention, a charge pump-type voltage doubler using large capacitors outside the chip is driven by an oscillation signal with a swing of 0 to 2×V_INthat is generated by an interior circuit of the chip, so that a large current can be output, and an MOS device is destroyed by the electrostatic discharge or latchup due to high voltage pins is avoided, thereby improving the efficiency of energy output of the circuit.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic diagram showing a circuit structure of a charge pump in the prior art;

FIG. 2 is a schematic diagram showing a circuit structure of another charge pump in the prior art;

FIG. 3 is a schematic diagram showing a circuit structure of a voltage doubler in the prior art;

FIG. 4 is a structural representation of a signal output apparatus according to Embodiment 1 of the invention;

FIG. 5 is a schematic diagram showing a circuit structure of a signal output apparatus according to Embodiment 1 of the invention;

FIG. 6 is a schematic diagram showing a circuit structure of a signal output apparatus according to Embodiment 2 of the invention;

FIG. 7 is a schematic diagram showing a circuit structure of a signal output apparatus according to Embodiment 3 of the invention;

FIG. 8 is a structural representation of a charge pump according to Embodiment 4 of the invention;

FIG. 9 is a schematic diagram showing a circuit structure of a charge pump according to Embodiment 5 of the invention;

FIG. 10 is a structural representation of a voltage doubler according to Embodiment 6 of the invention;

FIG. 11 is a schematic diagram showing a circuit structure of a voltage doubler according to Embodiment 7 of the invention;

FIG. 12 is a waveform timing diagram of each key node potential during one period in which a circuit works stably according to Embodiment 7 of the invention;

FIG. 13 is a schematic diagram showing a circuit structure of a voltage doubler according to Embodiment 8 of the invention;

FIG. 14 is a waveform timing diagram of each key node potential during one period in which a circuit works stably according to Embodiment 8 of the invention;

FIG. 15 is a schematic diagram showing a circuit structure of a voltage doubler according to Embodiment 9 of the invention;

FIG. 16 is a waveform timing diagram of each key node potential during one period in which a circuit works stably according to Embodiment 9 of the invention;

FIG. 17 is a schematic diagram showing a circuit structure of a voltage doubler according to Embodiment 10 of the invention; and

FIG. 18 is a waveform timing diagram of each key node potential during one period in which a circuit works stably according to Embodiment 10 of the invention.

DETAILED DESCRIPTION

The invention will now be described in detail in conjunction with the drawings.

FIG. 4 shows a structural representation of a signal output apparatus according to Embodiment 1 of the invention. The signal output apparatus is adapted to output a periodically varying signal with a swing in a range of 0 to 2×V_IN. As shown in the figure, the signal output apparatus includes: a first signal generating unit 11 and a second signal generating unit 12. The first signal generating unit 11 is adapted to generate a first oscillation signal with a swing of input voltage V_INto 2×V_IN, and a second oscillation signal with a swing of 0 to V_INand a phase the same as that of the first oscillation signal; the second signal generating unit 12 is adapted to take the first oscillation signal and the second oscillation signal as triggering signals, and convert the input voltage to a first output signal with a swing of 0 to 2×V_INand output the first output signal, where the phase of the first output signal is the same as that of the first oscillation signal.

Specifically, the second signal generating unit 12 includes various devices, taking an MOS transistor as an example, the second generating unit 12 may include a first PMOS and a first NMOS. FIG. 5 is a schematic diagram showing the circuit structure of the signal output apparatus according to Embodiment 1 of the invention, where the gate of the first PMOS is connected to an input voltage V_IN, and the source or the drain of the first PMOS is connected to the bulk and receives a first oscillation signal generated by the first signal generating unit 11; the source or the drain of the first NMOS receives a second oscillation signal generated by the first signal generating unit 11, the bulk of the first NMOS is grounded, and the gate of the first NMOS is connected to the input voltage V_IN; one of the source and the drain of the first PMOS that is not connected to the bulk is connected with one of the source and the drain of the first NMOS that does not receive the second oscillation signal, and outputs a first output signal at the connection point.

Additionally, the first signal generating unit 11 may also generate a third oscillation signal with an opposite phase but the same swing as the first oscillation signal, the function of which will be described in the subsequent embodiments.

The first signal generating unit 11 may be composed of various circuits. Two circuit structures of Embodiment 1 will now be provided by respectively taking Embodiment 2 and Embodiment 3 as an example, and the operating process thereof will be illustrated via the description of the circuit structures.

FIG. 6 shows a schematic diagram showing the circuit structure of the signal output apparatus according to Embodiment 2 of the invention. In the figure, P1 is the first PMOS, and N1 is the first NMOS.

The first signal generating unit 11 may generate one or more first oscillation signals with a swing of V_INto 2×V_INand second oscillation signals with a swing of 0 to V_INas desired, and the phases of these oscillation signals are optional. Because the unit does not need to output a large current, the capacitance in the unit may be small, thus an integrated capacitor inside the chip may be used, without occupying the chip pins. The operating process of the first signal generating unit 11 is as follows.

It is hypothesized that, initially, the voltage differences between the upper and lower plates of the capacitors C3 and C4 are both 0, the phases of the voltages output by the node L2 and the node R2 are opposite, and the two voltages are oscillation signals with a swing of 0 to V_IN. When the potential of the node L2 is 0 and the potential of the R2 is V_IN, the voltage differences between the upper and lower plates of the capacitors C4 and C3 are still kept at 0 volt due to the capacitance characteristic, thus initially, the potential of the node L is 0 volt, and the potential of the node R is raised to V_IN. Then the N5 is conducting, the capacitor C3 is charged by the input voltage V_IN, and the potential of the node L is pulled up to V_IN. When the potential of the node L2 is inverted to V_INand the potential of the node R2 is inverted to 0, because the voltage differences between the upper and lower plates of the capacitors C3 and C4 can not be changed at the moment the oscillation signals are inverted, at the moment after inversion, the potential of the node L is pushed up to a voltage of about 2×V_INdue to the ascending of the potential of the node L2, and the potential of the node R is pulled down to 0 due to the descending of the potential of the node R2. In such a case, the N5 is not conducting, while the N6 is conducting because the potential of the gate L is higher than the potential of the node R, thus the potential of the node R is also pulled up to V_IN. Thereafter, according to the above principle, the voltages of the nodes L and R are opposite phase voltages, and the potentials thereof vary periodically in a range of V_INto about 2×V_INrespectively; the voltages of the nodes L2 and R2 are also opposite phase voltages, and the potentials thereof vary periodically in a range of 0 to V_INrespectively. In this embodiment, the oscillation signals output by the nodes R and R2 are taken as the first oscillation signal and the second oscillation signal respectively, and the oscillation signal output by the node L is taken as the third oscillation signal.

The voltage output by the node L2 is obtained via the P4 and N4 shown in FIG. 6, the gate of P4 is connected to a clock signal ck2w, and the gate of N4 receives a clock signal ck1n; the voltage output by the node R2 is obtained via the P3 and N3 shown in FIG. 6, the gate of P3 is connected to a clock signal ck1w, and the gate of N3 receives a clock signal ck2n, where the ck1w and the ck2w are two-phase non-overlap oscillation signals, the ck1n and the ck2n are two-phase non-overlap oscillation signals, the ck2n and the ck2w are phase-inverted oscillation signals, and the ck1n and the ck1w are phase-inverted oscillation signals.

The role of the second signal generating unit 12 in FIG. 6 is to generate an oscillation signal with a swing of 0 to about 2×V_INby using the first oscillation signal output by the output node R of the first signal generating unit 11 and the second oscillation signal output by the output node R2. A NMOS transistor N1 and a PMOS transistor P1 exist in the circuit structure of the second signal generating unit 12, the first oscillation signal from the node R is output to the source and the bulk of P1, the second oscillation signal from the node R2 is output to the source of N1, the gates of P1 and N1 are connected to the input voltage V_IN, the bulk of N1 is grounded, the drain of P1 is connected with the drain of N1, and an output signal with a swing of 0 to about 2×V_INis output. The specific operating process thereof is as follows.

When the potential of the R2 is 0 and the potential of the R is V_IN, the N1 is conducting, while the P1 is blocking, and the potential of the first output signal (labelled as hclk1 in this embodiment) is pulled down to 0 by the N1; when the potential of the R2 is V_INand the potential of the R is about 2×V_IN, N1 is blocking while P1 is conducting, and the potential of the hclk1 is pulled up to about 2×V_INby the P1. Thereafter, according to the above principle, the hclk1 oscillates in a range of 0 to about 2×V_IN.

Embodiment 3 of the invention further provides a schematic diagram showing the circuit structure of another signal output apparatus. As shown in FIG. 7, P1 in the figure is the first PMOS, and N1 is the first NMOS. The operating process of the first signal generating unit 11 is as follows: it is hypothesized that, initially, the voltage differences between the upper and lower plates of the capacitors C3 and C4 are both 0, and nodes ck1w and ck2w are two-phase non-overlap clock signals with opposite phases and a swing of 0 volt to V_INvolt. To simplify the description of the operating principle, the high level dead zone between the potentials of the nodes ck1w and ck2w is ignored (that is, it is hypothesized that Tb−Ta=0), and such a simplification will not influence the main operating principle. When the potential of the node ck1w is 0, the potential of the ck2w is V_IN. The voltage differences between the upper and lower plates of the capacitors C4 and C3 are still kept at 0 due to the capacitance characteristic, thus the potential of the node L is 0, while the potential of the node R is raised to V_IN. Then, the P13 is conducting, the capacitor C3 is charged by the input voltage V_IN, and the potential of the node L is pulled up to about V_IN. Then, when the potential of the node ck1w is inverted to V_INand the potential of the node ck2w is inverted to 0, because the voltage differences between the upper and lower plates of the capacitors C3 and C4 can not be changed at the moment the clock signals are inverted, at the moment after inversion, the potential of the node L is pushed up to the potential of about 2×V_INdue to the ascending of the potential of the node ck1w, and the potential of the node R is pulled down to 0 due to the descending of the potential of the node ck2w. In such a case, the P13 is not conducting, while the P14 is conducting because the potential of the gate R is lower than V_IN, thus the capacitor C4 is charged by the input voltage V_IN, and the potential of the node R is also pulled up to about V_IN. Thereafter, according to the above principle, the potentials of the nodes L and R will vary periodically in a range of V_INto about 2×V_INrespectively. In this embodiment, the signal output by the node R is taken as the first oscillation signal, the signal output by the node ck2w is taken as the second oscillation signal, and the oscillation signal output by the node L is taken as the third oscillation signal.

The operating principle of the second signal generating unit 12 is the same as that in Embodiment 2. When the potential of the ck2w is 0 and the potential of the R is V_IN, the N1 is conducting, while the P1 is blocking, the potential of the first output signal (labelled as hclk1 in this embodiment) is pulled down to 0 by the N1; when the potential of the ck2w is V_INand the potential of the R is about 2×V_IN, the N1 is blocking while the P1 is conducting, and the potential of the hclk1 is pulled up to about 2×V_INby the P1. Thereafter, according to the above principle, hclk1 oscillates in a range of 0 to about 2×V_IN.

According to the description of Embodiment 1 to Embodiment 3, a signal output apparatus for independently generating an oscillation signal with a swing of 0 to 2×V_INis provided. The signal output apparatus provided in these embodiments has a broad application range, for example, it can be used in an analog digital converter (ADC), an apparatus for converting a low voltage digital signal to a high voltage digital signal, a charge pump and a voltage doubler.

FIG. 8 shows a structural representation of a charge pump according to Embodiment 4 of the invention. The charge pump includes a signal output module 21 and a charge storage module 22, where the signal output module 21 is adapted to output a first output signal with a swing of 0 to 2×V_IN, the signal output module 21 may be the signal output apparatus described in Embodiment 1 to Embodiment 3, and may also have other circuit structures; the charge storage module 22 is adapted to take the first output signal as a triggering signal to generate two states, where one state is to charge one plate of the capacitor by the input voltage, and the other state is to provide a voltage higher than 0 volt to the other plate of the capacitor.

It is hypothesized that the signal output module 21 is the signal output apparatus in Embodiment 2, as shown in FIG. 9. FIG. 9 is a schematic diagram showing the circuit structure of the charge pump according to Embodiment 5 of the invention. The charge storage module 22 includes a ninth PMOS (P9 in FIG. 9) and a first capacitor (C1 in FIG. 9), where the gate of the ninth PMOS receives the first output signal, one of the source and the drain of the ninth PMOS is connected to the input voltage V_IN, and the other of the source and the drain of the ninth PMOS is connected to the bulk of the ninth PMOS; the upper plate of the first capacitor is connected with one of the source and the drain of the ninth PMOS that is connected with the bulk, and the lower plate of the first capacitor receives a fourth oscillation signal with a swing of 0 to V_IN, where the phase of the fourth oscillation signal is the same as that of the first output signal.

In order to improve the efficiency of energy output of the charge pump, the fourth oscillation signal may also be obtained via the following structure.

The charge storage module 22 further includes a second PMOS (P2 in FIG. 9) and a second NMOS (N2 in FIG. 9), where the first clock signal ck1w is the input for the gate of the second PMOS, and the input voltage V_INis the input for the source and the bulk of the second PMOS; the second clock signal ck2n is the input for the gate of the second NMOS, the source and the bulk of the second NMOS are grounded, the drain of the second NMOS is connected with the drain of the second PMOS, and the connection point CS1 outputs the fourth oscillation signal.

The P9 charging the capacitor C1 in the charge storage module 22 uses the first output signal with a swing of 0 to about 2×V_INthat is output by the signal output apparatus, i.e., the hclk1. In a half period of charging, the potential of the node hclk1 is 0, P9 is conducting, the first capacitor C1 is charged by the input voltage V_IN, and the potential of the upper plate CP1 of C1 may reach up to V_IN; furthermore, because the potential of the lower plate CS1 of C1 is 0 at this point, the capacitor may be filled with charges up to the hilt.

In this embodiment, the lower plate of the C1 may be connected directly with the entity that provides a fourth oscillation signal with a swing of 0 to V_IN. For example, in FIG. 9, the P2 and the N2 constitute the entity that generates the fourth oscillation signal, where the P2 and the N2 are controlled by using two-phase non-overlap clock signals ck1w and ck2n, thus the breakover current from the input voltage V_INto the ground, which is generated due to the P2 and the N2 are conducting at the same time at the moment the clock signal level is inverted, may be eliminated, the power consumption may be reduced, and the efficiency of the energy output of the circuit may be improved. In the specific application of this embodiment, because the sizes of the MOS transistors P2 and N2 are larger than the sizes of the MOS transistors P4, N4, P3 and N3, the charge amount output by the output node CS1 is greater than the charge amount output by the node L2 or the node R2.

The charge pump in FIG. 9 further includes a first signal generating unit 11 and a second signal generating unit 12 in the signal output module 21.

The operating process of the first signal generating unit 11 is as follows.

It is hypothesized that, initially, the voltage differences between the upper and lower plates of the capacitors C3 and C4 are both 0, the phases of the voltages output by the node L2 and the node R2 are opposite, and the two voltages are oscillation signals with a swing of 0 to V_IN. When the potential of the node L2 is 0 and the potential of the R2 is V_IN, the voltage differences between the upper and lower plates of the capacitors C4 and C3 are still kept at 0 volt due to the capacitance characteristic, thus initially, the potential of the node L is 0 volt, and the potential of the node R is raised to V_IN. Then the N5 is conducting, the capacitor C3 is charged by the input voltage V_IN, and the potential of the node L is pulled up to V_IN. When the potential of the node L2 is inverted to V_INand the potential of the node R2 is inverted to 0, because the voltage differences between the upper and lower plates of the capacitors C3 and C4 can not be changed at the moment the oscillation signals are inverted, at the moment after inversion, the potential of the node L is pushed up to a voltage of about 2×V_INdue to the ascending of the potential of the node L2, and the potential of the node R is pulled down to 0 due to the descending of the potential of the node R2. In such a case, the N5 is not conducting, while the N6 is conducting because the potential of the gate L is higher than the potential of the node R, thus the potential of the node R is also pulled up to V_IN. Thereafter, according to the above principle, the voltages of the nodes L and R are opposite phase voltages, and the potentials thereof vary periodically in a range of V_INto about 2×V_INrespectively; the voltages of the nodes L2 and R2 are also opposite phase voltages, and the potentials thereof vary periodically in a range of 0 to V_INrespectively. In this embodiment, the oscillation signals output by the nodes R and R2 are taken as the first oscillation signal and the second oscillation signal, and the oscillation signal output by the node L is taken as the third oscillation signal.

The voltage output by the node L2 is obtained via the P4 and N4 in FIG. 9, the gate of P4 receives a clock signal ck2w, and the gate of N4 receives a clock signal ck1n; the voltage output by the node R2 is obtained via the P3 and N3 in FIG. 9, the gate of P3 receives a clock signal ck1w, and the gate of N3 receives a clock signal ck2n, where in the first signal generating unit 11, the ck1w and the ck2w are two-phase non-overlap oscillation signals, the ck1n and the ck2n are two-phase non-overlap oscillation signals, the ck2n and the ck2w are phase-inverted oscillation signals, and the ck1n and the ck1w are phase-inverted oscillation signals.

The specific operating process of the second signal generating unit 12 is as follows.

When the potential of the R2 is 0 and the potential of the R is V_IN, the N1 is conducting, while the P1 is blocking, and the potential of the first output signal is pulled down to 0 by the N1; when the potential of the R2 is V_INand the potential of the R is about 2×V_IN, the N1 is blocking while the P1 is conducting, and the potential of the hclk1 is pulled up to about 2×V_INby the P1. Thereafter, according to the above principle, the hclk1 oscillates in a range of 0 to about 2×V_IN.

Embodiment 6 of the invention further provides a voltage doubler, the structural representation of which is as shown in FIG. 10. The signal output module 31 in the figure may be the signal output apparatus described in Embodiment 1 to Embodiment 3, and may also be modules obtained by the other circuit connection modes that can output the first output signal with a swing of 0 to 2×V_INand the first oscillation signal and the third oscillation signal with a swing of V_INto 2×V_INbut with opposite phases, as in Embodiment 1 to Embodiment 3.

The charge storage module 32 in the voltage doubler may be the same as the charge storage module 22 of the charge pump in Embodiments 4 and 5, and may also be modules obtained by the other circuit connection modes that can take the first output signal as a triggering signal to generate two states, where one state is to charge one plate of the capacitor via the input voltage V_IN, and the other state is to provide a voltage higher than 0 volt to the other plate of the capacitor.

The voltage doubler further includes a current output switch module 33, which is adapted to take the first oscillation signal and the third oscillation signal as triggering signals to output the charges stored in the charge storage module 32 via a capacitor.

The voltage doubler in Embodiment 7 of the invention employs the signal output apparatus in Embodiment 2 and the charge storage module in Embodiment 5, and a schematic diagram of the circuit structure thereof is as shown in FIG. 11.

The voltage doubler in FIG. 11 includes a first signal generating unit 11 and a second signal generating unit 12 in the signal output module 31.

The operating process of the first signal generating unit 11 is as follows.

The voltage output by the node L2 is obtained via the P4 and the N4 in FIG. 11, the gate of the P4 receives a clock signal ck2w, and the gate of the N4 receives a clock signal ck1n; the voltage output by the node R2 is obtained via the P3 and the N3 in FIG. 11, the gate of the P3 receives a clock signal ck1w, and the gate of the N3 receives a clock signal ck2n, where in the first signal generating unit 11, the ck1w and the ck2w are two-phase non-overlap oscillation signals, the ck1n and the ck2n are two-phase non-overlap oscillation signals, the ck2n and the ck2w are phase-inverted oscillation signals, and the ck1n and the ck1w are phase-inverted oscillation signals.

The specific operating process of the second signal generating unit 12 is as follows.

The current output switch module 33 in the voltage doubler shown in FIG. 11 includes a tenth PMOS (P10 in FIG. 11), a fifth PMOS (P5 in FIG. 11) and a sixth PMOS (P6 in FIG. 11), where the gate of the tenth PMOS receives a first oscillation signal, one of the source and the drain of the tenth PMOS receives a third oscillation signal, the other one is connected with the bulk of the tenth PMOS, and one of the source and the drain of the tenth PMOS that is connected with the bulk is connected with the source or the drain of the fifth PMOS.

One of the source and the drain of the fifth PMOS that is not connected with the source or the drain of the tenth PMOS receives the first oscillation signal, the gate of the fifth PMOS receives the third oscillation signal, and the bulk of the fifth PMOS is connected with the bulks of the tenth PMOS and the sixth PMOS.

The gate of the sixth PMOS receives the third oscillation signal, the bulk of the sixth PMOS is connected with the bulk of the tenth PMOS, one of the source and the drain of the sixth PMOS is connected with the bulk of the ninth PMOS, and the other one of the source and the drain of the sixth PMOS is connected with the final output terminal, outputs an output voltage V_OUTthat is higher than the input voltage V_IN, and outputs a current.

The operating process of the charge storage module 32 and the current output switch module 33 is as follows.

The P9 charging the capacitor C1 in the charge storage module 32 uses the first output signal with a swing between 0 to about 2×V_INoutput by the signal output apparatus, i.e., hclk1 in FIG. 11. In a half period of charging, the potential of the node L is about 2×V_IN, the P6 in the current output switch module 33 is blocking; the potential of the node hclk1 is 0, the P9 is conducting, the first capacitor C1 is charged by the input voltage V_IN, and the potential of the upper plate CP1 of C1 may reach up to V_IN; furthermore, because the potential of the lower plate CS1 of C1 is 0 at this point, the capacitor may be filled with charges up to the hilt. Additionally, the lower plate of C1 may be connected directly with the entity that provides a fourth oscillation signal with a swing of 0 to V_IN, and in FIG. 11, the P2 and the N2 constitute the entity that generates the fourth oscillation signal, wherein the P2 and the N2 are controlled by using the two-phase non-overlap clock signals ck1w and ck2n.

However, in a half period of discharging, the potential of the lower plate CS1 of C1 is V_IN, because the voltage difference between the upper and lower plates of C1 can not be changed at the moment the oscillation signal is inverted, at the moment after inversion, the upper plate CP1 of C1 is pushed up to a voltage of about 2×V_INdue to the ascending of the potential of the lower plate CS1; while at this point, the potential of the node L is about V_IN, P6 in the current output switch module 33 is conducting, and the V_OUTis discharged by the C1 via the connection of a load capacitor and a load resistor. Thus it can be seen that, after an operation time long enough, the voltage of the V_OUTmay be stabilized at about twice of the input voltage V_IN, and a large current may be output.

The waveform timing of each key node potential during one period in which a circuit works stably according to Embodiment 7 is as shown in FIG. 12, where a dead zone exists between Ta and Tb, a half period of discharging exists between Tb and Tc, and a half period of charging exists between Tc and Td. The swing and phase of the signal that is output by each node will now be respectively described.

The swings of the ck1w, the ck2w, the ck1n and the ck2n are O-V_IN, and the ck1w and the ck2w are two-phase non-overlap oscillation signals, the ck1n and the ck2n are two-phase non-overlap oscillation signals, the ck2n and the ck2w are phase-inverted oscillation signals, and the ck1n and the ck1w are phase-inverted oscillation signals. The signals output by the nodes L2 and R2 have a swing of 0 to V_INand opposite phases. In the half period of discharging, the potential of the L2 is 0 and the potential of the R2 is V_IN; in the half period of charging, the potential of the L2 is V_INand the potential of the R2 is 0. The signals output by the nodes L and R have a swing of V_INto 2×V_INand opposite phases. In the half period of discharging, the potential of the L is V_INand the potential of the R is 2×V_IN; in the half period of charging, the potential of the L is 2×V_INand the potential of the R is V_IN. The signal output by the node hclk1 has a swing of 0 to 2×V_INand the phase of the signal output by the node hclk1 is the same as that of the signal output by the node R. In the half period of discharging, the potential of the hclk1 is 2×V_IN; in the half period of charging, the potential of the hclk1 is 0. The phases of the signals output by the node CS1 and the node CP1 are the same, the swing of the signal output by the CS1 is 0 to V_IN, and the swing of the signal output by the CP1 is V_INto 2×V_IN. In the half period of discharging, the potential of the CS1 is V_IN, and the potential of the CP1 is 2×V_IN; in the half period of charging, the potential of the CS1 is 0 and the potential of the CP1 is V_IN.

Additionally, in the current output switch module 33 of this embodiment, an oscillation signal with a higher level is alternately selected, by using the switches P5 and P10, from a third oscillation signal and a first oscillation signal with a swing of V_INto about 2×V_INrespectively output from the node L and the node R, to output to the bulks of the P6, P5 and P10, thus ensuring that the well potentials of the P6, P5 and P10 can be kept at about 2×V_IN. Furthermore, the charge pump according to this embodiment is applied in the situation of the large current output, and the size of the output switch P6 is relatively large, the parasitic capacitance between the N well and the P-type substrate of the output switch P6 is sufficient to keep the well potential of the output switch P6 at about 2×V_IN, thus no additional capacitors need to be added between the bulks of the P6, P5 and P10 and the ground.

The voltage doubler in Embodiment 7 can only output a large current in one half period, and cannot output a current in the other half period. In order to further increase the efficiency of energy output, an improvement is made on the voltage doubler of Embodiment 7. FIG. 13 shows a schematic diagram showing the circuit structure of the voltage doubler according to Embodiment 8 of the invention. The signal output apparatus in this embodiment is similar to that in Embodiment 7, except that the two second signal generating units in Embodiment 2 are combined to obtain a first output signal and a second output signal with a swing of 0˜2×V_IN(corresponding to the two oscillation signals with opposite phases output by the nodes hclk1 and hclk2 in FIG. 13).

The first signal generating unit 11 generates a fifth oscillation signal with a swing of 0 to V_IN(the signal output by the node L2 in FIG. 13), and the phases of the fifth oscillation signal and the second oscillation signal are opposite; the second signal generating unit 12 is further adapted to trigger a plurality of self-contained devices connected with the input voltage to be conducting or blocking by using the third oscillation signal (the signal output by the node L in FIG. 13) and the fifth oscillation signal, and to output a second output signal with a swing of 0 to 2×V_IN, where the phases of the second output signal and the first output signal are opposite; the charge storage module 32 further includes: a seventh PMOS (P7 in FIG. 13) and a second capacitor (C2 in FIG. 13), where the gate of the seventh PMOS receives the second output signal, one of the source and the drain of the seventh PMOS is connected to the input voltage V_IN, and the other of the source and the drain is connected to the bulk; one plate of the second capacitor is connected with one of the source and the drain of the seventh PMOS that is connected with the bulk, and the other plate receives a sixth oscillation signal with a swing of 0 to V_IN, where the phases of the sixth oscillation signal and the second output signal are the same; the current output switch module 33 further includes: an eighth PMOS (P8 in FIG. 13), of which the gate receives the first oscillation signal, and the bulk is connected with the bulk of the sixth PMOS (P6 in FIG. 13); and one of the source and the drain of the eighth PMOS is connected with the bulk of the seventh PMOS (P7 in FIG. 13), and the other of the source and the drain is connected with the output terminal and outputs charges.

The operating process of the first signal generating unit 11 is as follows.

It is hypothesized that, initially, the voltage differences between the upper and lower plates of the capacitors C3 and C4 are both 0, the phases of the voltages output by the node L2 and the node R2 are opposite, and the two voltages are oscillation signals with a swing of 0 to V_IN. When the potential of the node L2 is 0 and the potential of the R2 is V_IN, the voltage differences between the upper and lower plates of the capacitors C4 and C3 are still kept at 0 volt due to the capacitance characteristic, thus initially the potential of the node L is 0 volt, and the potential of the node R is raised to V_IN. Then the N5 is conducting, the capacitor C3 is charged by the input voltage V_IN, and the potential of the node L is pulled up to V_IN. When the potential of the node L2 is inverted to V_INand the potential of the node R2 is inverted to 0, because the voltage differences between the upper and lower plates of the capacitors C3 and C4 can not be changed at the moment the oscillation signals are inverted, at the moment after inversion, the potential of the node L is pushed up to a voltage of about 2×V_INdue to the ascending of the potential of the node L2, and the potential of the node R is pulled down to 0 due to the descending of the potential of the node R2. In such a case, the N5 is not conducting, while the N6 is conducting because the potential of the gate L is higher than the potential of the node R, thus the potential of the node R is also pulled up to V_IN. Thereafter, according to the above principle, the voltages of the nodes L and R are opposite phase voltages, and the potentials thereof vary periodically in a range of V_INto about 2×V_INrespectively; the voltages of the nodes L2 and R2 are also opposite phase voltages, and the potentials thereof vary periodically in a range of 0 to V_INrespectively. In this embodiment, the oscillation signals output by the nodes R and R2 are taken as the first oscillation signal and the second oscillation signal respectively, and the oscillation signal output by the node L is taken as the third oscillation signal.

The specific operating process of the second signal generating unit 12 is as follows.

When the potential of the R2 is 0 and the potential of the R is V_IN, the N1 is conducting, while the P1 is blocking, and the potential of the first output signal (hclk1 in FIG. 13) is pulled down to 0 by the N1; when the potential of the R2 is V_INand the potential of the R is about 2×V_IN, the N1 is blocking while the P1 is conducting, and the potential of the hclk1 is pulled up to about 2×V_INby the P1. Thereafter, according to the above principle, the hclk1 oscillates in a range of 0 to about 2×V_IN. By the same token, the oscillation signals output by the node L and the node L2 is output to the P12 and the N12, and the hclk2 is output.

To eliminate the breakover current caused by the simultaneous conducting of metal oxides, and to furthest decrease the power consumption and improve the efficiency of current output, the first signal generating unit generates a second oscillation signal and a fifth oscillation signal by using two-phase non-overlap oscillation signals. Specifically, the first signal generating unit 11 includes: a third PMOS (P3 in FIG. 13), a third NMOS (N3 in FIG. 13), a fourth PMOS (P4 in FIG. 13) and a fourth NMOS (N4 in FIG. 13).

The gate of the third PMOS P3 receives a first clock signal (the ck1w in the first signal generating unit 11 of FIG. 13), and one of the source and the drain receives the input voltage V_INand is connected to the bulk; the gate of the third NMOS N3 receives a second clock signal (the ck2n in the first signal generating unit 11 of FIG. 13), and one of the source and the drain is connected to the bulk; one of the source and the drain of the third PMOS P3 that is not connected with the bulk is connected with one of the source and the drain of the third NMOS N3 that is not connected with the bulk, and the second oscillation signal of the connection point (R2 in FIG. 13) is output.

The gate of the fourth PMOS P4 receives a third clock signal (the ck2w in the first signal generating unit 11 of FIG. 13), and one of the source and the drain receives the input voltage V_INand is connected to the bulk; the gate of the fourth NMOS N4 receives a fourth clock signal (the ck1n in the first signal generating unit 11 of FIG. 13), and one of the source and the drain is connected to the bulk; one of the source and the drain of the fourth PMOS P4 is connected with one of the source and the drain of the fourth NMOS N4 that is not connected with the bulk, and the fifth oscillation signal of the connection point (L2 in FIG. 13) is output.

Where the ck1w and the ck2w are two-phase non-overlap oscillation signals, the ck1n and the ck2n are two-phase non-overlap oscillation signals, the ck2n and the ck2w are phase-inverted oscillation signals, and the ck1n and the ck1w are phase-inverted oscillation signals.

In the charge storage module 32, the node CS1 and the node CS2 respectively output a fourth oscillation signal and a sixth oscillation signal with a swing of 0 to V_INand opposite phases, and the fourth oscillation signal output by the node CS1 is obtained in the following way: the gate of the P2 receives the clock signal ck1w, and one of the source and the drain receives the V_IN; the gate of the N2 receives the clock signal ck2n, and one of the source and the drain is connected with the bulk and is grounded; one of the source and the drain of the P2 that does not receive the V_INis connected with one of the source and the drain of the N2 that is not connected with the bulk, and the connection point CS1 outputs the fourth oscillation signal. The sixth oscillation signal output by the node CS2 is obtained in the following way: the gate of the eleventh PMOS (P11 in FIG. 13) receives the clock signal ck2w, one of the source and the drain receives the V_IN; the gate of the eleventh NMOS (N11 in FIG. 13) receives the clock signal ck1n, and one of the source and the drain is connected with the bulk and is grounded; one of the source and the drain of the P11 that does not receive the V_INis connected with one of the source and the drain of the N11 that is not connected with the bulk, and the connection point CS2 outputs the sixth oscillation signal. Where the ck1w and the ck2w are two-phase non-overlap oscillation signals, the ck1n and the ck2n are two-phase non-overlap oscillation signals, the ck2n and the ck2w are phase-inverted oscillation signals, and the ck1n and the ck1w are phase-inverted oscillation signals.

The operating process of the circuit of the voltage doubler is as follows.

During the half period in which the potential of the node hclk1 is 0, the switch P9 is conducting, the first capacitor C1 is charged by the input voltage V_IN, the potential of the node L is 2×V_IN, and the current output switch P6 is blocking; at this point, the potential of the node hclk2 is 2×V_IN, the potential of the upper plate CP2 of the second capacitor C2 is 2×V_IN, the potential of the node R is V_IN, the current output switch P8 is conducting, and a large current is output to the node V_OUTby the C2 via connecting a load capacitor and a load resistor. During the half period in which the potential of the node hclk1 is 2×V_IN, the potential of the upper plate CP1 of C1 is 2×V_IN, the potential of the node L is V_IN, the current output switch P6 is conducting, and a large current is output to the node V_OUTby the C1 via connecting a load capacitor and a load resistor; while at this point, the potential of the node hclk2 is 0, the switch P7 is conducting, the C2 is charged by the input voltage V_IN, the potential of the node R is 2×V_IN, and the current output switch P8 is blocking. Additionally, an oscillation signal with a higher level is alternately selected, by using the switches P5 and P10, from a third oscillation signal and a first oscillation signal with a swing of V_INto about 2×V_INrespectively output from the node L and the node R, and is output to the bulks of the P6, P8, P5 and P10, thus ensuring that the well potentials of the P6, P8, P5 and P10 is kept at about 2×V_IN.

The waveform timing of each key node potential during one period in which a circuit works stably according to Embodiment 8 is as shown in FIG. 14. In this way, a large current is output during each half period; in comparison with Embodiment 7, the charge amount output in one period is doubled, and the output current is also approximately doubled. In FIG. 14, a dead zone exists between Ta and Tb, a half period during which the C1 discharges (a half period during which the C2 is charged) exists between Tb and Tc, and a half period during which the C1 is charged (a half period during which the C2 discharges) exists between Tc and Td. The swing and phase of the signal that is output by each node will now be respectively described.

The swing of the ck1w, the ck2w, the ck1n and the ck2n is 0-V_IN, and the ck1w and the ck2w are two-phase non-overlap oscillation signals, the ck1n and the ck2n are two-phase non-overlap oscillation signals, the ck2n and the ck2w are phase-inverted oscillation signals, and the ck1n and the ck1w are phase-inverted oscillation signals. The signals output by the nodes L2 and R2 have a swing of 0 to V_INand opposite phases, and in the half period during which the C1 discharges (i.e., the half period during which the C2 is charged), the potential of the L2 is 0 and the potential of the R2 is V_IN; in the half period during which the C1 is charged (i.e., the half period during which the C2 discharges), the potential of the L2 is V_INand the potential of the R2 is 0. The signals output by the nodes L and R have a swing of V_INto 2×V_INand opposite phases, and in the half period during which the C1 discharges (i.e., the half period during which the C2 is charged), the potential of the L is V_INand the potential of the R is 2×V_IN; in the half period during which the C1 is charged (i.e., the half period during which the C2 discharges), the potential of the L is 2×V_INand the potential of the R is V_IN. The signals output by the node hclk1 and the node hclk2 have a swing of 0 to 2×V_INand opposite phases, and in the half period during which the C1 discharges (i.e., the half period during which the C2 is charged), the potential of the hclk1 is 2×V_INand the potential of the hclk2 is 0; in the half period during which the C1 is charged (i.e., the half period during which the C2 discharges), the potential of the hclk1 is 0 and the potential of the hclk1 is 2×V_IN. The signals output by the node CS1 and the node CS2 have a swing of 0 to V_INand opposite phases, and in the half period during which the C1 discharges (i.e., the half period during which the C2 is charged), the potential of the CS1 is V_INand the potential of the CS2 is 0; in the half period during which the C1 is charged (i.e., the half period during which the C2 discharges), the potential of the CS1 is 0 and the potential of the CS2 is V_IN. The signals output by the node CP1 and the node CP2 have a swing of V_INto 2×V_INand opposite phases, and in the half period during which the C1 discharges (i.e., the half period during which the C2 is charged), the potential of the CP1 is 2×V_INand the potential of the CP2 is V_IN; in the half period during which the C1 is charged (i.e., the half period during which the C2 discharges), the potential of the CP1 is V_INand the potential of the CP2 is 2×V_IN.

FIG. 15 shows a schematic diagram showing the circuit structure of the voltage doubler according to Embodiment 9 of the invention. It is set that the signal output module 31 is the signal output apparatus in Embodiment 3, and that the charge storage module 32 and the current output switch module 33 are the same as those in Embodiment 7.

The operating process of the first signal generating unit 11 in the figure is as follows: it is hypothesized that, initially, the voltage differences between the upper and lower plates of the capacitors C3 and C4 are both 0, and the nodes ck1w and ck2w are two-phase non-overlap clock signals with opposite phases and a swing of 0 volt to V_INvolt. To simplify the description of the operating principle, the high level dead zone between the potentials of the nodes ck1w and ck2w is ignored (i.e., it is hypothesized that Tb−Ta=0), and such a simplification will not influence the main operating principle. When the potential of the node ck1w is 0, the potential of the ck2w is V_IN. The voltage differences between the upper and lower plates of the capacitors C4 and C3 are still kept at 0 due to the capacitance characteristic, thus the potential of the node L is 0, while the potential of the node R is raised to V_IN. Then the P13 is conducting, the capacitor C3 is charged by the input voltage V_IN, and the potential of the node L is pulled up to about V_IN. Then, when the potential of the node ck1w is inverted to V_INwhile the potential of the node ck2w is inverted to 0, because the voltage differences between the upper and lower plates of the capacitors C3 and C4 can not be changed at the moment the clock signals are inverted, at the moment after inversion, the potential of the node L is pushed up to the potential of about 2×V_INdue to the ascending of the potential of the node ck1w, and the potential of the node R is pulled down to 0 due to the descending of the potential of the node ck2w. In such a case, the P13 is not conducting, while the P14 is conducting because the potential of the gate R is lower than V_IN, thus the capacitor C4 is charged by the input voltage V_IN, and the potential of the node R is also pulled up to about V_IN. Thereafter, according to the above principle, the potentials of the nodes L and R vary periodically in a range of V_INto about 2×V_INrespectively. In this embodiment, the signal output by the node R is taken as the first oscillation signal, the signal output by the ck2w is taken as the second oscillation signal, and the oscillation signal output by the node L is taken as the third oscillation signal.

The operating principle of the second signal generating unit 12 is as follows: when the potential of the ck2w is 0 and the potential of the R is V_IN, the N1 is conducting while the P1 is blocking, and the potential of the first output signal (labelled as hclk1 in this embodiment) is pulled down to 0 by the N1; when the potential of the ck2w is V_INand the potential of the R is about 2×V_IN, the N1 is blocking while the P1 is conducting, and the potential of the hclk1 is pulled up to about 2×V_INby the P1. Thereafter, according to the above principle, the hclk1 oscillates in a range of 0 to about 2×V_IN.

The operating principle of the charge storage module 32 is as follows.

The P9 that charges the capacitor C1 uses the first output signal with a swing of 0 to about 2×V_IN, i.e., the hclk1. In the half period of charging, the potential of the node hclk1 is 0, the P9 is conducting, the first capacitor C1 is charged by the input voltage V_IN, and the potential of the upper plate CP1 of C1 may reach up to V_IN; furthermore, because the potential of the lower plate CS1 of C1 is 0 at this point, the capacitor may be filled with charges up to the hilt.

In this embodiment, the lower plate of the C1 may be connected directly with the entity that provides a fourth oscillation signal with a swing of 0 to V_IN. For example, in FIG. 15, the P2 and the N2 constitute the entity that generates the fourth oscillation signal, where the P2 and the N2 are controlled using the two-phase non-overlap clock signals ck1w and ck2n, thus the breakover current from the input voltage V_INto the ground, which is generated at the moment the clock signal level is inverted because the P2 and the N2 are conducting at the same time, may be eliminated, the power consumption may be reduced, and the efficiency of energy output of the circuit may be improved.

The operating principle of the current output switch module 33 is as follows.

In the half period of charging, the potential of the node L is about 2×V_IN, and the P6 in the current output switch module 33 is blocking; the potential of the node hclk1 is 0, the P9 is conducting, the first capacitor C1 is charged by the input voltage V_IN, and the potential of the upper plate CP1 of C1 may reach up to V_IN; furthermore, because the potential of the lower plate CS1 of C1 is 0 at this point, the capacitor may be filled with charges up to the hilt. Additionally, the lower plate of the C1 may be connected directly with the entity that provides a fourth oscillation signal with a swing of 0 to V_IN, in FIG. 15, the P2 and the N2 constitute the entity that generates the fourth oscillation signal, where the P2 and the N2 are controlled via the two-phase non-overlap clock signals ck1w and ck2n.

However, in the half period of discharging, the potential of the lower plate CS1 of C1 is V_IN, and because the voltage difference between the upper and lower plates of the C1 can not be changed at the moment the oscillation signal is inverted, at the moment after inversion, the upper plate CP1 of the C1 is pushed up to a voltage of about 2×V_INdue to the ascending of the potential of the lower plate CS1; while at this point, the potential of the node L is about V_IN, the P6 in the current output switch module 33 is conducting, and charges are output via the C1 by connecting a load capacitor and a load resistor. Thus it can be seen that, after an operation time long enough, the voltage of the V_OUTmay be stabilized at about twice of the input voltage V_IN, and a large current may be output.

Additionally, an oscillation signal with a higher level is alternately selected, by using the switches P5 and P10, from a third oscillation signal and a first oscillation signal with a swing of V_INto about 2×V_INrespectively output from the node L and the node R, and is output to the bulks of the P6, P5 and P10, thus ensuring that the well potentials of the P6, P5 and P10 can be kept at about 2×V_IN.

FIG. 16 is a waveform timing diagram of each key node potential during one period in which a circuit works stably according to Embodiment 9. A dead zone exists between Ta and Tb, a half period of discharging exists between Tb and Tc′, a dead zone exists between Tc′ and Tc, a half period of charging exists between Tc and Td′, and a dead zone exists between Td′ and Td. The swing and phase of the signal that is output by each node will now be respectively described.

In this embodiment, because the nodes L and R are respectively connected to the two-phase non-overlap clock signals ck1w and ck2w via the lower plates of the capacitors C3 and C4, the nodes L and R also output two-phase non-overlap oscillation signals. Such a structure has the following three advantages.

1) Because the charge storage module 32 requires a large current output capacity, the size of the driver of the lower plate of the capacitor in the charge storage module 32 is relatively large. After the two-phase non-overlap clock signals are used, the P2 and the N2 that have large sizes may be blocking simultaneously in the dead zone (between the time Ta and the time Tb as shown in FIG. 16), so that the breakover current from the input voltage V_INto the ground caused by the simultaneous conducting of the P2 and the N2, which have large sizes, may be avoided.

2) Because the signals output by the nodes L and R are two-phase non-overlap oscillation signals with a swing of V_INto 2×V_IN, a high level dead zone (between the time Ta and the time Tb as shown in FIG. 16) exists when the two signals oscillate, thus the backflow of charges from the N well of the P6, P5 and P10 to the nodes L and R may be avoided, and this is favorable to keep the high potential of the N well, the conducting of the PN junction between the source/drain of the P6 and the N well of the P6 may be avoided, and thereby reducing the charges leaked to the N well from the CP1 at a high potential, and reducing the charge loss of the charge storage module 32.

3) Because the node hclk1 reaches a high level before the node CP1 reaches a high level (at the time Ta and Tb as shown in FIG. 16), the P9 is blocking, so that the charges flow back from the node CP1 to the V_INmay be avoided, thus the charge loss of the charge storage module 32 may be reduced.

As similar to the voltage doubler in Embodiment 7, the voltage doubler in Embodiment 9 can only output a large current in one half period, and cannot output a current in the other half period. In order to further increase the output efficiency, an improvement is made on the charge pump in Embodiment 9. FIG. 17 shows a schematic diagram showing the circuit structure of the voltage doubler according to Embodiment 10 of the invention. In this embodiment, two second signal generating units 12 are combined, and a first output signal and a second output signal with a swing of 0 to 2×V_IN(i.e., the two two-phase non-overlap phase-inverted oscillation signals output by the nodes hclk1 and hclk2 in FIG. 17) are obtained.

The operating process of the first signal generating unit 11 in the figure is as follows.

It is hypothesized that, initially, the voltage differences between the upper and lower plates of the capacitors C3 and C4 are both 0, and the nodes ck1w and ck2w are two-phase non-overlap clock signals with opposite phases and a swing of 0 volt to V_INvolt. When the potential of the node ck1w is 0, the potential of the ck2w is V_IN. The voltage differences between the upper and lower plates of the capacitors C4 and C3 are still kept at 0 due to the capacitance characteristic, thus the potential of the node L is 0, while the potential of the node R is raised to V_IN. Then the P13 is conducting, the capacitor C3 is charged by the input voltage V_IN, and the potential of the node L is pulled up to about V_IN. Then, when the potential of the node ck1w is inverted to V_INand the potential of the node ck2w is inverted to 0, because the voltage differences between the upper and lower plates of the capacitors C3 and C4 can not be changed at the moment the clock signals are inverted, at the moment after inversion, the potential of the node L is pushed up to the potential of about 2×V_INdue to the ascending of the potential of the node ck1w, and the potential of the node R is pulled down to 0 due to the descending of the potential of the node ck2w. In such a case, the P13 is not conducting, while the P14 is conducting because the potential of the gate R is lower than V_IN, thus the capacitor C4 is charged by the input voltage V_IN, and the potential of the node R is also pulled up to about V_IN. Thereafter, according to the above principle, the potentials of the nodes L and R periodically vary in a range of V_INto about 2×V_INrespectively. In this embodiment, the signal output by the node R is taken as the first oscillation signal, the signal output by the ck2w is taken as the second oscillation signal, the signal output by the ck1w is taken as the fifth oscillation signal, and the oscillation signal output by the node L is taken as the third oscillation signal.

The operating principle of the second signal generating unit 12 is as follows.

When the potential of the ck2w is 0 and the potential of the R is V_IN, the N1 is conducting, while the P1 is blocking, and the potential of the first output signal (labelled as hclk1 in this embodiment) is pulled down to 0 by the N1; when the potential of the ck2w is V_INand the potential of the R is about 2×V_IN, the N1 is blocking while the P1 is conducting, and the potential of the hclk1 is pulled up to about 2×V_INby the P1. Thereafter, according to the above principle, the hclk1 oscillates in a range of 0 to about 2×V_IN. By the same token, the oscillation signals output by the nodes L and ck1w are output to the P12 and the N12, and the hclk2 is output.

The operating principle of the charge storage module 32 is as follows.

However, in the half period of discharging, the potential of the lower plate CS1 of C1 is V_IN, because the voltage difference between the upper and lower plates of the C1 can not be changed at the moment the oscillation signal is inverted, at the moment after inversion, the upper plate CP1 of the C1 is pushed up to a voltage of about 2×V_INdue to the ascending of the potential of the lower plate CS1. By the same token, the signal of the upper plate CP2 of the C2 is output via the cooperation of the hclk2 and the P7.

In this embodiment, the lower plate of the C1 may be connected directly with the entity that provides a fourth oscillation signal with a swing of 0 to V_IN. For example, in FIG. 17, the P2 and the N2 constitute the entity that generates the fourth oscillation signal, where the P2 and the N2 are controlled via the two-phase non-overlap clock signals ck1w and ck2n. By the same token, an entity constituted by the P11 and the N11 outputs a sixth oscillation signal at the node C52, where the P11 and the N11 are controlled by using the two-phase non-overlap clock signals ck2w and ck1n.

The operating principle of the current output switch module 33 is as follows.

During the half period in which the potential of the node hclk1 is 0, the switch P9 is conducting, the first capacitor C1 is charged by the input voltage V_IN, the potential of the node L is 2×V_IN, and the current output switch P6 is blocking; at this point, the potential of the node hclk2 is 2×V_IN, the potential of the upper plate CP2 of the second capacitor C2 is 2×V_IN, the potential of the node R is V_IN, the current output switch P8 is conducting, and a large current is output to the node V_OUTby the C2 via connecting a load capacitor and a load resistor. During the half period in which the potential of the node hclk1 is 2×V_IN, the potential of the upper plate CP1 of the C1 is 2×V_IN, the potential of the node L is V_IN, the current output switch P6 is conducting, and a large current is output to the node V_OUTvia the C1 by connecting a load capacitor and a load resistor; while at this point, the potential of the node hclk2 is 0, the switch P7 is conducting, the C2 is charged by the input voltage V_IN, the potential of the node R is 2×V_IN, and the current output switch P8 is blocking. Additionally, an oscillation signal with a higher level is alternately selected, by using the switches P5 and P10, from a third oscillation signal and a first oscillation signal with a swing of V_INto about 2×V_INrespectively output from the node L and the node R, and is output to the bulks of the P6, P8, P5 and P10, thus ensuring that the well potentials of the P6, P8, P5 and P10 are kept at about 2×V_IN.

The waveform timing of each key node potential during one period in which a circuit works stably according to Embodiment 10 is as shown in FIG. 18. A dead zone exists between Ta and Tb, a half period during which the C1 discharges (a half period during which the C2 is charged) exists between Tb and Tc′, a dead zone exists between Tc′ and Tc, a half period during which the C1 is charged (a half period during which the C2 discharges) exists between Tc and Td′, and a dead zone exists between Td′ and Td. The swing and phase of the signal that is output by each node will now be respectively described.

The swing of the ck1w, the ck2w, the ck1n and the ck2n is 0 to V_IN, and the ck1w and the ck2w are two-phase non-overlap oscillation signals, the ck1n and the ck2n are two-phase non-overlap oscillation signals, the ck2n and the ck2w are phase-inverted oscillation signals, and the ck1n and the ck1w are phase-inverted oscillation signals. The signals output by the nodes L and R have a swing of V_INto 2×V_INand opposite phases. In the half period during which the C1 discharges (i.e., the half period during which the C2 is charged), the potential of the L is V_INand the potential of the R is 2×V_IN; and in the half period during which the C1 is charged (i.e., the half period during which the C2 discharges), the potential of the L is 2×V_INand the potential of the R is V_IN. The signals output by the node hclk1 and the node hclk2 have a swing of 0 to 2×V_INand opposite phases. In the half period during which the C1 discharges (i.e., the half period during which the C2 is charged), the potential of the hclk1 is 2×V_INand the potential of the hclk2 is 0; in the half period during which the C1 is charged (i.e., the half period during which the C2 discharges), the potential of the hclk1 is 0 and the potential of the hclk2 is 2×V_IN. The signals output by the node CS1 and the node CS2 have a swing of 0 to V_INand opposite phases. In the half period during which the C1 discharges (i.e., the half period during which the C2 is charged), the potential of the CS1 is V_INand the potential of the CS2 is 0; and in the half period during which the C1 is charged (i.e., the half period during which the C2 discharges), the potential of the CS1 is 0 and the potential of the CS2 is V_IN. The signals output by the node CP1 and the node CP2 have a swing of V_INto 2×V_INand opposite phases. In the half period during which the C1 discharges (i.e., the half period during which the C2 is charged), the potential of the CP1 is 2×V_INand the potential of the CP2 is V_IN; in the half period during which the C1 is charged (i.e., the half period during which the C2 discharges), the potential of the CP1 is V_INand the potential of the CP2 is 2×V_IN.

Thus, a large current is output during each half period. In comparison with Embodiment 9, the charge amount output in one period is doubled, and the output current is also approximately doubled.

In addition to obtaining the same effects as Embodiment 9, Embodiment 10 further has the following advantages.

Because the voltages of the signals output by the nodes L and R have a swing of V_INto 2×V_INand opposite phases and the two signals are two-phase non-overlap oscillation signals, a high level dead zone (between the time Ta and the time Tb as shown in FIG. 18) exists when the two signals oscillate, thus the problem that the charges flow back from the CP1 (or the CP2) with a high potential to the CP2 (or the CP1) with a low potential via the V_OUTand the well shared by the P6, P8, P5 and P10 due to the simultaneous conducting of the switches P6 and P8 and the switches P5 and P10 when the ordinary inverted clock signal is inverted may be solved.

According to the above description of each embodiment of the invention, Embodiment 11 of the invention further provides a method for outputting a large current, which includes the following processes:

Process 1: generating three oscillation signals, where the first oscillation signal has a swing of V_INto 2×V_IN, the second oscillation signal has a swing of 0 to V_IN, the phases of the first oscillation signal and the second oscillation signal are the same, and the third oscillation signal and the first oscillation signal have the same swing and opposite phases;

Process 2: taking the first oscillation signal and the second oscillation signal as triggering signals, converting the input voltage to a first output signal with a swing of 0 to 2×V_INand outputting the first output signal, where the phases of the first output signal and the first oscillation signal are the same;

Process 3: taking the first output signal as a triggering signal to generate two states, where one state is to charge one plate of a capacitor by an input voltage V_IN, and the other state is to provide a voltage higher than 0 volt to the other plate of the capacitor; and

Process 4: taking the first oscillation signal and the third oscillation signal as triggering signals to output the charges on the plate of the capacitor.

Finally, it should be noted that, in each embodiment of the invention, the circuit structure is described by taking the MOS (for example, MOS field effect transistor (MOSFET) and Enhanced MOSFET) as an example. Additionally, other types of transistors, such as Bipolar Transistor, may also be used in the circuit provided in the invention. Moreover, the objects of the invention may also be attained by using a plurality of diodes and triodes put in circuit.

By using the signal output apparatus, the charge pump, the voltage doubler and the method for outputting a current according to the embodiments of the invention, under the premise that a standard CMOS manufacturing process is used and the voltage differences born by the lead wires between the gate and the source, the source and the drain and the gate and the drain of all the MOS devices should not exceed V_INwhen a charge pump-type voltage doubler works stably, it may be ensured that the MOS is not fragile to be destroyed by the electrostatic discharge or latchup introduced via each pin of the chip, and at the same time, the waste on the charges output by the charge pump may be reduced; additionally, because the two-phase non-overlap oscillation signals are used, the charge loss may be avoided to the maximum extent, and the efficiency of energy output of the circuit may be improved.

It will be appreciated that one skilled in the art may make various modifications and alterations to the present invention without departing from the spirit and scope of the present invention. Accordingly, if these modifications and alterations to the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention intends to include all these modifications and alterations.

Signal Output Apparatus, a Charge Pump, a Voltage Doubler and a Method to Output Current

Information

Publication Number

Date Filed

Date Published

Inventors

Original Assignees

CPC

US Classifications

International Classifications

Abstract

Description

Claims

Priority Claims (1)

PCT Information