This invention relates to systems and methods for producing bidirectional power conversion. More specifically, this invention relates to circuits and methods for creating a truly bidirectional power converter with multiple control loops and power routing.
As the need for smaller and more intricate integrated circuits surfaces, so does the need for smaller power routing and power conversion processes. Power routing and power conversion processes may be found, for example, in uninterruptible power systems, next-generation automotive electrical systems, redundant battery systems for computers, telecommunications equipment and portable electronic devices.
Bidirectional power conversion provides the ability to change power attributes to and from a component. Bidirectional power conversion may be found in portable devices such as cell phones and Personal Digital Assistants (PDAs) that contain rechargeable batteries. These DC rechargeable batteries are automatically charged when connected to an AC power source. During battery charging, the power from the AC power source is converted into DC power (either externally with a wall adapter or within the portable device circuitry), and the battery is charged. Additionally, during battery charging, the portable device operates using the power supplied from the AC power source. When the portable device is not connected to the AC power source or wall adapter, the battery serves as the primary power source. This process of supplying power to and receiving power from the battery is in accordance with the principles of bidirectional power conversion. Furthermore, the process of routing power from the DC battery instead of the AC power source or wall adapter is present in such systems.
In prior systems, truly bidirectional power conversion is not realized. Instead, prior systems perform bidirectional power conversion through two separate unidirectional power converters and an additional power routing system to switch between the two converters. Each separate unidirectional power converter is dedicated to the power conversion process associated with a particular direction. The result of using multiple power converters in addition to additional logic for the purpose of routing in prior systems is that the amount of switches present is inopportunely large.
In light of the foregoing, it would be desirable to provide a truly bidirectional power converter with multiple control loops and power routing in order to reduce the complexity, size and cost associated with current converters that utilize multiple unidirectional power converters.
The present invention provides systems and methods for providing bidirectional power conversion using a single converter and controller for power routing. Through the use of only a single converter and controller, the amount of switches used in fabricating the circuit of the present invention is significantly reduced. In fact, the number of necessary switches may be nearly halved. Moreover, the volume, cost, complexity, and size of a bidirectional power converter in accordance with the principles of the present invention may also be nearly halved.
In addition to the above, further consideration relating to bidirectional power conversion should preferably also be given to the need for relatively high efficiency, which is measured as a ratio of the power output to the power input. Higher efficiency can be achieved by reducing system losses, and can enable extended battery operating time. The present invention therefore combines power routing and power conversion, and in doing so, eliminates the necessity for several switches and reduces total power (I2R) losses. Furthermore, the power supply switching techniques used in accordance with the principles of the present invention provide additional efficiency in bidirectional power conversion systems.
Moreover, although the following methods and systems to be presented assume external AC-DC-conversion (e.g., power from an AC power source is converted externally by a wall adapter to supply DC power to be used in the bidirectional power converter), the present invention is not limited in this manner. AC-DC and DC-AC converters are readily available and may be incorporated into the embodiments of the present invention, though they have been excluded from the embodiments described in applicants' specification in order to avoid the inclusion of material that would unnecessarily complicate the description of the present invention.
The above-mentioned objects and features of the present invention will be more clearly understood from the following detailed description considered in conjunction with the following drawings, in which the same reference numerals denote the same structural elements throughout, and in which:
Switch S0131 of
Reactive elements 120 may be one or more components in which power may be supplied to (and stored) as well as taken from. For example, reactive elements 120 may be any combination of inductors, transformers or capacitors. The invention is not limited in this manner.
In accordance with the principles of the present invention, switches S0131 to SN 133 may perform up to four functions. As explained in more detail below, these functions include charging reactive elements 120, providing synchronous rectification, discharging reactive elements 120 (to drop any stored charge), and providing power routing.
Moreover, although switches are shown to couple each of the power buses to the reactive elements in FIG. 1 and the figures to follow, other types of circuit components which are suitable for a particular application may be used. For example, assuming that power bus B1142 in
Controller 110 in
In one example, bus B0141 can be a DC wall adapter, bus B1142 can be a battery, and bus BN 143 can be a portable electronic device. It should be noted that although the use of a wall adapter results in external conversion of power from an AC power source to DC power, such conversion may be incorporated directly into the operation of a bidirectional power converter in accordance with the principles of the present invention. Accordingly, B0141 can be an AC power source, in which case some type of rectification would be included within the bidirectional power converter utilizing the AC power source. In the example above, controller 110 may be a current-mode controller that uses multiple control loops in order to OPEN and CLOSE the switches, and thereby control the transferring of power among the buses. As explained in greater detail below, controller 110 controls the flow of power among the various buses by adjusting the magnitude and direction of current flow through reactive elements 120, from and to the various power buses in the circuit.
Additionally, it should be understood by those skilled in the art that not every switch shown in
Circuit 200 includes controller 210, buses B0241 and B1242, switches S1231 and SGND 232, and inductor 220. The circuit shown in
In bidirectional converter 200, because only a simple inductor 220 is used as the reactive element, the device may be fabricated with relatively small size and high density. Moreover, bidirectional converter 200 works well for bus voltages V0<<V1 (V0 is the voltage at bus B0241 and V1 is the voltage at bus B1242). As V0 approaches V1, however, the duty factor required for operating as a boost converter may become a problem. This problem may be alleviated, as explained, below, through the use of a flyback transformer instead of inductor 220.
Similar to controller 110 of
For the circuit shown in
In addition, persons of ordinary skill will appreciate that, assuming the normal voltage requirements for a buck or boost converter are met, bus B0341 shown in
One manner in which the above power transfer combination problem can be avoided is to incorporate a switch between bus B0341 and inductor 220. In this case, it is preferable to use a single-pole, double-throw switch. Furthermore, additional buses may be coupled to inductor 220 on the side coupled to bus B0341 without departing from the scope of the invention, and these additional buses also may or may not be separated from inductor 220 by switches.
It should be appreciated by those skilled in the art that the bidirectional buck-boost converter shown in FIG. 2 and described above is a non-insulating converter. This type of converter is generally used in applications where voltage is only required to be stepped up or down by a relatively small amount, and in which it is acceptable for there not to exist electrical isolation between the input and output of the converter. In certain situations, however, it becomes desirable or even necessary to obtain isolation between the converter input and output. This isolation may be accomplished by removing the inductor and incorporating a transformer into the converter. Specifically, flyback converters are a version of buck-boost converters that incorporate transformer isolation while storing energy.
Although flyback transformer 420 of circuit 400 is represented using the same symbol as an ideal transformer, it is better described as a “two-winding inductor.” Flyback transformers, unlike ideal transformers, do not allow current to flow simultaneously in both windings. Instead, the magnetizing inductance of a flyback transformer assumes the role of the inductor of the buck-boost converter, and the magnetizing current is switched between the primary and secondary windings.
Furthermore, low-side N-channel MOSFET switching, or ground-side switching, simplifies the switch driver design in circuit 400. In this case, the MOSFETs acting as switches have their respective gates connected to a controller, their drain terminals connected to the flyback transformer and their source terminals connected to ground via a sense circuit.
The main advantage of using N-channel MOSFETs results from the typically greater efficiency when compared to using a comparable on-resistance P-channel device (because less gate charge will be required for switching). Additionally, low-side switching with N-channel MOSFETs permits the use of MOSFETs with relatively low absolute maximum gate-source ratings, even in higher voltage applications. To further simplify the circuit design, common current sensing may be used to reduce the component count. For example, a sense resistor may be used, such as sense resistor 412 shown in FIG. 4. Alternatively, another suitable component may be used for this purpose.
An important consideration when dealing with converter 400 is the coupling coefficient of flyback transformer 420. The windings of a flyback transformer are said to be coupled when the are arranged in such a manner that a changing magnetic field created by one of the windings can induce a current in the other winding. Moreover, the coupling coefficient of transformer 420 indicates the degree of coupling between the windings, and should be made as high as possible (i.e., the coupling coefficient should be as close to one as possible).
The coupling coefficient of transformer 420 is dependent on the portion of the total flux lines that cuts both the primary and secondary windings. Preferably, all the flux lines generated by the primary winding should cut the secondary winding, and vice versa. Lines of flux generated by one winding that do not link with the secondary winding are referred to as leakage flux, and result in having a voltage induced in the secondary winding that is less than would otherwise occur under ideal conditions. For this reason, efficiency is reduced as a result of leakage, and spikes may result at the drain of the switching device.
Bifilar winding can be used to reduce leakage inductance and improve the magnetic coupling of the windings, and therefore, is often used when it is important to achieve a superior coupling coefficient and to improve overall efficiency. In this situation, a simple integer turns ratio may be desirable. Nevertheless, the turns ratio of the windings of flyback transformer 420 can be adjusted, although the resulting duty factor should be taken into consideration. Furthermore, particularly in low voltage applications, consideration should be given to selecting a transformer with low winding resistance in order to improve efficiency at heavier loads.
It should be understood that the ratio of V1 (the voltage at power bus B1442) to V0 (the voltage at power bus B0441) in
A particular benefit of the bidirectional converter shown in
A similar flyback converter topology to that shown in
Additional DC connections are made to flyback transformer 520 by adding multiple secondary windings. As shown, on either side of each winding, a switch and its corresponding bus voltage is connected. A person skilled in the art will appreciate that the polarity dots on the transformer are correct in situations where the voltages on all buses have the same polarity, for example, V0-VN are all positive or V0-VN are all negative. Moreover, the preferred power transfer combinations involve power transfer between bus B0541 and the remainder of the buses, but in some instances other combinations are also acceptable. Once again, as with the circuits described above, Isense bus 511 may convey one or more sense signals to controller 510. Assuming that only one power bus is receiving power from or supplying power to flyback transformer 520, a single current sense signal can be used (where the sense signal is sampled at certain time intervals). Alternatively, multiple sense signals can be used when more than one power bus is receiving or supplying power at any given time.
It should be understood by those skilled in the art that each possible power flow combination of the aforementioned bidirectional power converters may have a unique application, unique power converter topology, unique demands on transient response and unique input and output voltages or currents. For example, a bidirectional flyback converter may be operating in one of two different modes, namely, discontinuous conduction mode or continuous conduction mode. In the former, all energy stored in the first winding during the ON (energy storing) time is completely delivered to the other winding and to the load before the next cycle, and there may also be dead time between the instant the secondary current reaches zero and the start of the next cycle. A circuit operating in discontinuous mode is typically designed to enter into continuous mode as the output current level is increased beyond a predetermined value. When continuous mode is entered into, there remains energy in the secondary winding at the beginning of the next cycle. Furthermore, because of the many different requirements for different power flow combinations, each power transfer combination may require different feedback signals and compensation (examples of which are provided below), except that a sense current is typically required in order to determine the direction of current flow.
Depending on the desired power flow combination (i.e., the source(s) of power and the destination(s) of the power being transferred), decoder 610 selects the sense signals and compensation (feedback network) needed. In the embodiments discussed below, a direction signal serves to select the appropriate sense signal and feedback network. Moreover, in accordance with the principles of the present invention, and as shown in
As shown in
When a converter using the circuit shown in
Alternatively, when the converter using circuit 700 is operating as a regulator, error amplifier 740 uses the signal provided by the voltage sense line (supplied through an input network comprising a voltage divider made up of resistors R4764 and R5765) and uses a feedback network comprising resistor R6766 and capacitor C2772. In this manner, when circuit 700 is being used by a converter operating as a regulator, error amplifier 740 acts as an integrator with zero with the signal and feedback network described above. Using an integrator with zero configuration when the converter is operating as a regulator, as opposed to a simple integrator configuration as in the case above when the converter is operating as a charger, provides a faster transient response. Although having a faster transient response is particularly important when the converter is operating as a regulator as opposed to a battery charger, the integrator with zero configuration (or another suitable configuration) may be used in all modes of operation of the converter.
Moreover, for purposes of simplification, it is assumed that in circuit 700, error amplifier 740 can drive both feedback networks (i.e., the integrator and integrator with zero feedback networks). Thus, an output selection switch is not necessary as shown in
Oscillator 710, current amplifier 720, pulse-width modulation (PWM) comparator 730 (which receives a current ramp from oscillator 710) and PWM latch 750 provide the remaining components for circuit 700. A driver, not shown in
The circuit of
Control circuit 801 operates in the following manner. If the circuit in
The direction signal emerging from direction comparator 845, which measures the ability of wall adapter 810 to supply power, is fed into driver 860. Additionally, the direction signal controls switch SW1893, thereby selecting the input network and feedback network to be used by error amplifier 843 in receiving the sense signal. As explained above, an separate output switch is not necessary to select the appropriate feedback network, but may easily be implemented. Moreover, for battery charging, SW1893 preferably ensures that error amplifier 843 acts as an integrator (i.e., the feedback network includes a capacitor), while during voltage regulation it preferably ensures that error amplifier 843 acts as an integrator with zero (i.e., the feedback network includes a resistor in series with a capacitor), resulting in a faster transient response.
At the beginning of each oscillator cycle, PWM latch 872 is set. Depending on the desired direction of power flow, the appropriate switch is turned on. The remaining components that make up control circuit 801 for this converter ensure that the current flowing through inductor 890 remains below a predetermined level while power is being transferred among the circuit components.
A consideration relating to compensation in converter 800 is that the poles and zeros of the power circuit are different for the circuit when operated as a buck converter as opposed to a boost converter, the result being that the circuit shown in
In circuit 1000, voltage selection network 1064 selects the power source as either battery B11020 or wall adapter 1010 based on the voltages available and creates an output voltage for all the remaining devices. One typical scheme uses the voltage of the two available voltages that is greater in magnitude to create voltage VMAX 1067. Another technique would be to take the lower of the two available voltages from battery B11020 and wall adapter 1010, as long as it exceeds the minimum voltage needed by the remaining devices. This technique may provide added efficiency but also increases the complexity of voltage selection network 1064.
VDRIVE 1072 is an optional regulated voltage linearly regulated down from VMAX 1076 which may be needed when power MOSFETs M11062 and M21063 have low absolute maximum gate-source voltages. In low voltage applications, Schottky diodes 1068 and 1069 are placed in parallel (as shown in
It should be understood by those skilled in the art that a regulated wall adapter may be used in place of unregulated wall adapter 1010, thereby eliminating the necessity for low dropout regulator 1084. In that case, the output to the circuitry and other devices (e.g., a portable electronic device) could be taken directly across capacitor C21051 shown in FIG. 10. The invention is not limited in this manner.
During the operation of power converter 1000, the direction signal emerging from comparator 1031 determines whether converter 1000 operates as a battery charger or voltage regulator. Based on this direction signal, switch SW21092 selects the appropriate sense signal, input network, and feedback network for error amplifier 1034 for either battery charge current control, or regulator output voltage control. Once again, an output switch is not used to select the appropriate feedback signal, but the present invention is not limited in this respect. The direction signal originating from comparator 1031 also feeds driver 1070 and determines whether M11062 charges (stores energy in) transformer T11061 and M21063 discharges T11061, or whether these roles for M11062 and M21063 are reversed.
During battery charging, switch SW21092 is in the down position, and resistors R71047 and R81046, along with capacitor C41054, create a simple integrator with error amplifier 1034 and also set up nominal charge current. At the start of the oscillator cycle, PWM latch 1082 is set, causing M21063 to turn on. When the switch current reaches a predetermined level, as set by the control voltage at the output of error amplifier 1034, M21063 turns off and M11062 turns on. Additionally, when transistor M21063 is off and a battery charging current is flowing through resistor R41044, switch SW11099 samples the current sense signal. In this case, battery charging current can flow either through the body diode or from drain to source of transistor M11062. For this reason, RDRIVE is the signal to use for observing the charging current.
When circuit 1000 is being used for voltage regulation (i.e., power is being supplied by the battery), switch SW21092 is in the up position and resistor R61048 and capacitor C31053 create an integrator with zero with error amplifier 1034, while resistors R11042 and R21043 set up the nominal output voltage of the regulator to be compared with voltage reference level 1083.
During voltage regulation, M11062 turns on at the start of each oscillator cycle. Moreover, when the transformer current reaches a predetermined level as set by the output of error amplifier 1034, M11062 turns off and M21063 turns on. Current amplifier 1033 increases the amplitude of the current sense signal (observed as a voltage drop across resistor R41044) while rejecting noise arising from M11062 and M21063 gate charge, switching transients from primary and secondary currents of transformer T11061, common-mode noise potentially from other circuits, and ringing that may result from imperfect coupling between transformer windings. As shown in
When transformer T11061 has fully dumped its energy, the MOSFET that was serving as the synchronous rectifier may be shut OFF. This is accomplished through the use of zero current comparator 1032, which shuts OFF either M11062 or M21063 at the appropriate time. This is commonly done when power converter 1000 is expected to enter discontinuous conduction at light loading in order to improve the efficiency at lighter loads.
Comparator 1035 serves as a PWM comparator for peak-current mode control power converter 1000. Oscillator 1081 triggers the start of a power conversion cycle by setting PWM latch 1082 (driving Q high), assuming the reset-dominant input of PWM latch 1082 is not high. This causes the MOSFET (either M11062 or M21063 depending on the direction of power transfer) to turn on, thereby charging transformer T11061. As the energy stored in transformer T11061 increases, the current through sense resistor R41044 also increases, and, therefore, the voltage at the non-inverting input of PWM comparator 1035 also increases. When the voltage at the non-inverting input of PWM comparator 1035 exceeds the voltage on the inverting input, the output of PWM comparator 1035 goes high, resetting PWM latch 1082. As a result, the MOSFET that was charging transformer T11061 is turned off. In addition, it should be noted that oscillator 1081 provides PWM comparator 1035 a ramp current (similar to oscillator 710 described above). However, a voltage ramp could also be used with minor circuit modifications. The same is true for each of the bidirectional power converter circuits described in accordance with the principles of the present invention.
At this point, neither of MOSFETs M11062 or M21063 are turned on, and transformer T11061 begins to discharge through the body diode of the opposite MOSFET. This is a result of the flyback action and polarity of the windings of the transformer. After the break time (i.e., the time that MOSFETs M11062 and M21063 are off) has elapsed, the opposite MOSFET turns on connecting transformer T11061 to the load side of the circuit. If at any time during this part of the power conversion cycle the inductor current reaches zero, zero current comparator 1032 shuts off the MOSFET that was connecting transformer T11061 to the load side of the circuit. The circuit then waits for oscillator 1081 to provide a pulse to the S input of PWM latch 1082. When this pulse is provided to the S input of PWM latch 1082, another complete power conversion cycle begins.
Capacitors C11052 and C21051 are included at least for the purpose of filtering the current pulses front the power converter, reducing electromagnetic interference, ripple voltage, and ripple current. Moreover, the sizing of capacitors C11052 and C21051, which preferably have low effective series resistance and inductance, should be selected depending on desired circuit characteristics (e.g., ripple voltage amount).
It should be noted that transformer saturation, which typically can occur at high output currents and extreme duty cycles, should be avoided under all operating conditions and combinations. In addition, although not shown, duty cycle limiting and pulse skipping at light loading is generally recommended for this topology. For example, Burst Mode™ of operation (hereinafter, “Burst Mode”) may be entered into in order to provide higher operating efficiency. Examples of regulators that use Burst Mode are Linear Technology Corp.'s 1435 and 1735 series products.
When operating in Burst Mode, the power MOSFETs operate intermittently at light loads, thereby increasing efficiency by minimizing switching losses. A Burst comparator may be used to determine when Burst Mode should be enabled, during which time the MOSFETs are not switched according to their normal cycles. It also determines when Burst Mode should be disabled, at which time normal circuit operation is resumed. Due to the introduction of broadband noise during Burst Mode operation, in particular during battery charging, it may be desirable to use normal operation.
In bidirectional converter 1100, direction comparator 1131 determines if a wall adapter is connected and operating (i.e., supplying sufficient power) by sensing the voltage drop across diode D11185. This requires the voltage supplied by wall adapter 1010 to be several hundred millivolts higher (at room temperature) than the voltage the bidirectional converter supplies in order to adequately forward bias diode D11185. This reduces net efficiency (wall adapter to battery) slightly.
Smaller differential voltages (wall adapter to circuit) may be used to sense a powered wall adapter by using resistive wall adapter detection. This is accomplished by replacing diode D11185 with a current sense resistor (not shown).
There are, however, several disadvantages to using resistive wall adapter detection. First, the drop across the current sense resistor would vary with the current supplied by wall adapter 1010. Although in a current-limited battery charger this may not be a problem until the float voltage is reached, the drop in the sense resistor nonetheless should be carefully chosen. Second, offset in wall adapter detection comparator 1031 would become more critical, as would the common mode rejection ratio, filtering, and circuit layout (Kelvin connections to the current sense resistor, etc.) Third, R31041 would load the power converter when power is being supplied by battery 1020, and thus, resistor R31041 would have to be made relatively large in order for it not to impact the efficiency of the converter. One possible solution to this problem is to replace R31041 with R11042 and R21043 by moving R11042 and R21043 to the right of the current sense resistor (and eliminating R31041). Alternatively, the diode-based wall adapter detection circuit shown in
The one embodiment of driver circuitry, like driver circuitry 1200, corresponds to component 860 in
Identical circuits 1251 and 1252 in
For each of circuits 1251 and 1252, the components therein are configured such that the two internal MOSFETs are precluded from being simultaneously ON. This is accomplished in circuit 1251 through the use of feedback signals 1271 and 1272, which respectively drive NAND gates 1282 and 1281. Similarly, in circuit 1252, feedback signals 1273 and 1274 drive NAND gates 1284 and 1283, respectively. Accordingly, in circuit 1251, N-channel MOSFET 1254 does not turn ON before P-channel MOSFET 1255 has turned OFF, and vice versa. The same situation is present for internal MOSFETs 1255 and 1256 of circuit 1252. Moreover, feedback signals 1275 and 1276, which drive NOR gates 1237 and 1236, respectively, are responsible for preventing RDRIVE 1261 and BDRIVE 1262 from turning the external power MOSFETs ON at the same time.
Level shifter circuits 1241-1243 shown in
Moreover, the circuit shown in
Store signal 1214 is used to determine when energy is stored in the reactive elements (i.e., when power from any of the available power buses capable of supplying power will be used to store energy in the windings of the transformer or in the inductor acting as the reactive element in the bidirectional power converter). In other words, when store signal 1214 is a logic high, this input triggers the start of a new power conversion cycle, and power is then transferred to the reactive elements to be stored. When store signal 1214 is a logic low, however, the energy previously stored in the reactive elements is transferred to any combination of the available power buses.
In the converters of
Synchronous rectification, or rectifier input 1212, on the other hand, receives a signal from a zero current comparator (component 841 in circuits 800 and 900 and component 1032 in circuits 1000 and 1100). This signal allows the driver to facilitate rectification using the switches within the respective circuits, thereby providing control over the power transfer.
In order to control the source of power for charging the reactive elements (e.g., either the battery or the wall adapter for the case of the circuits described above), driver circuitry 1200 also includes a selection signal input. Selection signal 1211 indicates whether or not a wall adapter is supplying power by sensing the presence or absence of a voltage drop across a diode. This signal is provided to driver circuitry 1200, and also selects the sense signal for the error amplifier found in the bidirectional power converter circuitry. For example, in
Finally, in the embodiment shown in
Further consideration when dealing with circuits that use current-mode controlling (e.g., applicants' claimed bidirectional power converter) should preferably be given to the use of slope compensation. In particular, when dealing with the switching in a fixed frequency, peak-current mode, pulse width modulated power supply, open loop instability or subharmonic oscillation can occur when the duty cycle (i.e., the ratio of the switch pulse width to the oscillator period) exceeds about 50%.
A typical solution to this problem is to use a slope compensation signal, derived from the oscillator, to ensure stability regardless of duty cycle. The slope compensation signal can be applied, for example, by either adding or subtracting a portion of the oscillator signal to the current sense signal. Moreover, this can be done by either adding or subtracting a portion of the oscillator signal to the non-inverting terminal of the PWM comparator or the inverting terminal of the PWM comparator.
The slope compensation signal described above causes the perceived rate of current rise in the reactive element to be greater than the rate of current fall, thereby allowing the regulator to operate at duty cycles greater than 50% without becoming unstable. In one embodiment, a slope compensation signal proportional to the duty cycle may be added to from the sense signal to increase the rate of current rise perceived by the control circuitry. As the slope compensation signal progresses towards its peak, the flow of current through the switch is impeded and an increase in the rate of current rise in the reactive element is perceived, allowing the converter to operate at duty cycles greater than 50% without losing stability.
Moreover, the reactive element current down slope (i.e., the current when the reactive element current is decreasing) is a function of the output voltage, or the output voltage minus the input voltage (depending on the power converter topology). For this reason, when the power converter voltages vary widely, the typical approach has been to add enough slope compensation to prevent instability even under the worst case operating conditions. Therefore, the result is excessive slope compensation being used under more normal operating conditions. This is particularly undesirable when dealing with bidirectional power converters, in which power converter voltages commonly have a large range of duty factors. Moreover, because the actual maximum current that can pass through the output switch is generally a function of the amount of slope compensation being used, the problem of overcompensation becomes especially detrimental in bidirectional power converters and causes the actual maximum current that can pass through the output switch to decrease proportionally as the duty cycle increases.
In particular, the slope compensation network shown in
Circuit 1300 shown in
Assuming conduct signal 1474 is a logic high, MOSFETs M11451 and M21452 select the correct gain control signal from either VBAT or VREG for the slope compensation (based on what voltage is determining the reactive elements downslope). These MOSFETs are turned ON and OFF, respectively, by NOR gates 1431 and 1432 (which are controlled by selection signal 1211). Furthermore, logic inverters 1411 and 1412 provides decoding for voltage selection and shutdown, and all remaining transistors and mosfets (labeled components 1453-1456, and 1461-1462 in
Circuit 1400, using the components described above and interconnected as shown in
The above described embodiments of the present invention are presented for purposes of illustration and not of limitation, and the present invention is limited only by the claims which follow.
Number | Name | Date | Kind |
---|---|---|---|
5256907 | Shimada et al. | Oct 1993 | A |
5541535 | Cao et al. | Jul 1996 | A |
5602464 | Linkowsky et al. | Feb 1997 | A |
5636108 | Taurand | Jun 1997 | A |
5682050 | Williams | Oct 1997 | A |
5731732 | Williams | Mar 1998 | A |
5734258 | Esser | Mar 1998 | A |
5745351 | Taurand | Apr 1998 | A |
5751150 | Rippel et al. | May 1998 | A |
5852559 | Li | Dec 1998 | A |
6121768 | Taurand | Sep 2000 | A |
6205035 | Vollmer et al. | Mar 2001 | B1 |
6384491 | O'Meara | May 2002 | B1 |