By

Jack Ivan J'maev

RELATED APPLICATIONS [0001] The present application claims priority, to such extent as allowed by law, to

United States Provisional Application Number 61/845,288 filed on July 11, 2013 and titled "METHOD AND SYSTEM FOR CONTINUOUS INDUCTOR CURRENT IN AC BUCK CONVERTERS" by J'maev; to United States Provisional Application Number 61/944,793 filed on February 26, 2014 and titled "METHOD AND SYSTEM FOR CONTINUOUS INDUCTOR CURRENT IN AC BUCK CONVERTERS" by J'maev ; and to United States National Phase Application 13/261,575 filed on January 26, 2013 and titled "METHOD AND APPARATUS FOR SYNCHRONOUS SINE WAVE DIMMING OF LUMINARIES" which is incorporated herein by reference. BACKGROUND

[0002] The notion of an Alternating Current (AC) Buck Converter appears to have been first described by Mason et al. in United States Patent 6,346,778 filed in January 1999. Several years later, Leah described several improvements to the AC Buck

Converter of Mason. In United States Patent 7,667,991 filed in July of 2007, Leah also described a method and apparatus for attenuating an AC input voltage according to a duty factor applied to a pulse width modulation controller.

[0003] Considering for the moment the teachings of Mason et al., a pulse width modulating reduction of an AC input results by using a main switch (or "buck switch") in conjunction with a commutation switch (or "freewheel switch") in order to reduce the AC input power by a duty factor applied to a pulse width modulation circuit. Mason et al. recognized that there was an issue with this fundamental concept when the output of the AC Buck Converter is directed to a reactive load. Mason et al. recognized that the polarity of the input voltage when compared to the polarity of the current flowing through the main switch could differ one relative to the other when a reactive load is driven by the AC Buck Converter. Mason et al. further recognized that when the polarity of the input voltage was different than the polarity of the current flowing through the main switch a loss of commutation followed and resulted in catastrophic failure of semiconductor switches that made up the main switch and the commutation switches.

[0004] The solution to this problem, according to Mason et al, provided for discontinuing pulse width modulation signaling to the main switch when the polarity of the current flowing through the main switch and the polarity of the input voltage were not the same. In one described embodiment, Mason et al. teaches that the control signal to the main switch should be extended so as to leave the main switch in an "ON" state when the voltage at the input to the AC Buck Converter and the current flowing through the main switch were of different polarities. We now realize that by leaving the main switch in the ON position during the time when the polarity of the input voltage and the current flowing through the main switch are different results in horrendous amounts of harmonic distortion in the output delivered to a reactive load.

[0005] Leah was not satisfied with the teachings of Mason et al. and advanced the art by providing a more continuous flow of power which did not require this extended ON state. According to Leah, it is possible to sense not only the polarity of the input voltage to an AC Buck Converter but also the polarity of current flowing to the load driven by said AC Buck converter. Based on this information, Leah taught us that the switching of the main switch and the switching of the commutation switch did not need to occur in a bidirectional manner. Rather, Leah taught us that, again based on the polarity of the input voltage and the polarity of the current flowing to the load, either the main switch or the commutation switch could be enabled in one direction and then enabled in both directions so as to ensure fully commutative pulse width modulation control. In Leah, switch sequencing occurred by first changing the state one switch in a pair of switches (for example the commutation switches) followed by a change of the state of a both switches in a second pair of switches (for example the main switch) then followed by a change of the state of the remaining switch in the first pair of switches. This resulted in two distinct intervals of time between switch state transitions.

[0006] Applicant realized that the teachings of Leah were in fact difficult to implement because it was simply difficult to obtain accurate information regarding the polarity of the input voltage and the polarity of the current flowing to the load. Any latency associated with these two input parameters relative to each other could result in a wrong switching decision and subsequent loss of commutation. Applicant realized that the polarity of the current flowing to the load and the polarity of the input voltage are in fact two distinct state variables that are non-synchronous relative to the switching decisions that need to be made in the embodiments of Leah.

[0007] Again furthering the art, Applicant disclosed a means by which all switches were controlled individually rather than in pairs as described by Leah. This resulted in a switch sequence where the state of a first switch was changed and then state of a second switch was changed and then the state of a third switch was changed and finally the state of the fourth switch was changed. This resulted in a switch sequence where there were three distinct intervals of time between switch state transitions. This latest teaching by Applicant, described in US Patent Application 13/261,575 (incorporated herein in its entirety by reference), relies only upon one state variable, that being the direction of current flow in an inductor that is included in an output filter in an AC Buck Converter. Because the switch sequencing is based upon a single state variable there is no potential for a wrong switching decision as in the teachings of Leah that require the use of two state variables. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Several alternative embodiments will hereinafter be described in conjunction with the appended drawings and figures, wherein like numerals denote like elements, and in which:

Fig. 1 is a block diagram that depicts an AC buck converter that uses three time intervals for switch sequencing;

Figs. 1A and IB are, respectively, a pictorial diagram and a timing diagram that depicts the operation of one illustrative example method and apparatus when current is flowing in a positive direction through a buck inductor;

Figs. 1C and ID are, respectively, a pictorial diagram and a timing diagram that depicts the operation of one illustrative example method and apparatus when current is flowing in a negative direction through a buck inductor;

Fig. 2 is a pictorial diagram that depicts the operation of four control signals controlling four switches in an AC buck converter;

Figs. 3 and 4 are pictorial diagrams that depict simulation results demonstrating loss of commutation;

Fig. 5A is a timing diagram that further clarifies current reversals during the termination of a freewheel cycle; Fig. 5B is a timing diagram that further clarifies current reversals during the termination of a buck cycle;

Fig. 6 shows one example method for preventing catastrophic failure by preventing current reversals during the termination of a buck cycle or the termination of a freewheel cycle based on sensing the level of current while in an extended state; Fig. 7 shows one alternative example method for preventing current reversals during the termination of a buck cycle or the termination of a freewheel cycle by predicting when current flow in an output inductor is not within a low current window; Fig. 8 is a flow diagram that depicts one example method for ensuring continuous current flow in an output inductor included in an AC buck converter;

Fig. 9 is a flow diagram that depicts one alternative example method for ensuring continuous current flow in an output inductor; Fig. 10 is a flow diagram that depicts a hybrid method for ensuring continuous current flow in an output inductor; and

Fig. 11 is a block diagram that depicts one example embodiment of a controller that ensures continuous conduction in an output inductor.

DETAILED DESCRIPTION

[0009] Fig. 1 is a block diagram that depicts an AC buck converter that uses three time intervals for switch sequencing. Although the present method and apparatus are described in the context of an AC buck converter that uses three time intervals for switch sequencing it is also directly applicable in an AC buck converter that that uses two time intervals for switch sequencing as described by Leah. In the context of an AC buck converter as described in US National Phase Application 13/261,575, the text and figures of which are incorporated into this disclosure in their entirety, there exists potential for catastrophic failure during switching of transistors in Fig.l when the current flowing through the inductor 360 transitions from positive to negative polarity. A similar failure can be experienced when the current flowing through the inductor 360 transitions from negative to positive polarity. As fully described in the incorporated reference, it is essential that current flowing through the inductor 360 continue without impairment. Should any discontinuity of current flow in the inductor 360 occur, then any back electromotive force in the inductor 360 results in a high- voltage spike that can damage either the freewheel transistors 370 and 375 and, to a lesser extent, the buck switch transistors 310 and 315.

[0010] Fig. 1, in essence, shows one example embodiment of an AC but converter that uses three intervals of time in sequencing of switches that are included in a buck switch and a freewheel switch. In this example embodiment, semiconductor devices are used for bi-directional switching of current in both the buck switch position 300 and the freewheel position 305. In this example method and embodiment, the buck switch 300 comprises two semiconductor switches, for example two MOSFETs. A first MOSFET 310 is disposed so as to receive the input AC waveform 320 at its drain terminal. The source of this first MOSFET 310 is electrically connected to a source terminal of a second

MOSFET 315 that is also included in the buck switch 300. The drain of this second MOSFET 315 comprises the output of the buck switch 300 and is connected to the buck inductor 360 and the freewheel switch 305. In this example method and embodiment, buck control is accomplished by MOSFET gate drivers 320, 325. It should be appreciated that these gate drivers comprise "high-side" drivers and, in other example embodiments, we include high-voltage isolation between the buck switch 300 and a control circuit 335.

[0011] In this example method and embodiment, the control circuit 335 operates relative (337) to the common terminal 340. Power for the control circuit 335 is derived from the input AC waveform directed (335) to the control circuit. This example embodiment also includes a buck inductor 360, the output of which is directed to an output terminal 365. This example method and embodiment further comprises a synchronous freewheel switch 305 comprising a third MOSFET 370 and a fourth

MOSFET 375. Third and fourth gate drivers 380 and 385 are also "high-side" drivers that are included in this example method and embodiment of the present art buck down- converter 390. In other alternative embodiments the "high- side" drivers are electrically isolated from the control circuit 335. It should be appreciated that even though this example embodiment is based on MOSFETs, any suitable switching mechanism may be used and the claims appended hereto are not intended to be limited to any particular type of switching mechanism.

[0012] Figs. 1A and IB are, respectively, a pictorial diagram and a timing diagram that depicts the operation of one illustrative example method and apparatus when current is flowing in a positive direction through a buck inductor. For the sake of convention, a current will be considered positive when it is flowing into a component, for example the buck inductor 1500. In this illustrative example method and apparatus, the synchronous freewheel comprises two synchronous freewheel switches identified as the positive freewheel "PF" 1520 and the negative freewheel "NF" 1525. In this illustrative method and apparatus, the freewheel switches comprise MOSFET devices disposed in a manner in which the drain of the negative freewheel MOSFET switch 1525 is electrically common to the buck inductor 1500 and buck switch. The positive freewheel MOSFET switch 1520 is disposed in a manner such that its drain terminal is connected to the common terminal of the apparatus. The MOSFETs 1520, 1525 forming the freewheel switch are dispose in a manner such that there source terminals are electrically common. This illustrative method and apparatus further comprise gate driving circuit 1530 and 1535 which enable control of the gate terminals of their respective MOSFETs. In this example method and apparatus, the buck switch comprises two switching devices 1540 and 1545. Accordingly, the buck switch comprises a positive buck switch "PB" 1540 and a negative buck switch "NB" 1545. Each buck switch is controlled by a gate drive circuit, depicted in the figure as 1550 and 1555. It should be appreciated that the scope of the claims appended hereto are not intended to be limited in scope to the use of MOSFET switches.

[0013] In order to maintain constant current flow through the buck inductor 1500, it becomes necessary to add an additional element to the apparatus and an additional step to the method supporting such apparatus. Mainly, an additional step includes sensing the direction of current flow in the buck inductor 1500, Accordingly, this illustrative apparatus further includes a current sensor 1560 which provides current sensing 1565 for the controller 1570. It should be appreciated that the structure of the buck switch in this alternative example method is analogous to the structure of the synchronous freewheel switch described above. In this illustrative example embodiment, the buck switch comprises the positive buck switch 1540 and a negative buck switch 1545 and each of these switches is disposed in parallel with an associated diode 1615 and 1620.

Accordingly, the positive buck diode 1620 is disposed with its cathode electrically common with the negative buck switch 1545 and is back-to-back with the negative buck diode 1615 wherein the negative buck diode 1615 is disposed in a manner such that its cathode is electrically common with the positive buck switch 1540. Although the positive buck switch 1540 and the negative buck switch 1545 are depicted in the figure as

MOSFETs, any suitable switch may be utilized, however positive and negative buck diodes (1620, 1615) must then be supplied in addition to the switches into switching devices do not include parasitic diodes analogous to that found in MOSFETs. As noted numerous times throughout this specification MOSFETs are a preferred device because of the parasitic diode is included in their structure. [0014] This alternative example method and apparatus are best understood through the teachings of the timing diagram (FIG IB) that depicts the sequence of switching the synchronous freewheel switches and the buck switches in the case where current is flowing into the buck inductor 1500 from the source 1580 or from the load 1585. In this figure, four control signals are depicted. These include a positive freewheel (PF) control signal 40, a negative freewheel (NF) control signal 50, a positive buck control signal 20 and a negative buck control signal 60. For the sake of clarity, these signals are depicted as "active high". Considering the situation when the synchronous freewheel switches are both engaged, positive current 1510 flows through the buck inductor 1500 because both the negative freewheel switch 1525 in the positive freewheel switch 1520 are both engaged by means of NF 50 and PF 40 control signals. This is illustrated by the timing diagram at point 1600 and it should be appreciated that while both freewheel switches are engaged, both buck switches are turned off (i.e. control signals PB 20 and NB 60 are low) as depicted in the timing diagram. According to this illustrative method and apparatus, switching from the freewheel state to the buck state occurs in a particular manner based upon the direction of current flowing through the buck inductor 1500. This current is flowing 1510 through the synchronous freewheel switches, the current must flow through the buck inductor 1500 because the buck switches are both turned off.

[0015] Transitioning to the buck state in this situation comprises a first step of disengaging 1605 the negative freewheel switch 1525. When the negative freewheel switch 1525 is disengaged 1605, it should be appreciated that current 1510 continues to flow through the buck inductor 1500 and is maintained by the negative freewheel diode 1610 which is disposed across the negative freewheel switch 1525 having its cathode electrically common with the buck switch and the inductor. Once the negative freewheel is disabled, it is now safe to engage 1630 the positive buck switch 1540. This allows current 1505 to flow from the source 1580 into the buck inductor 1500 even though the negative buck switch 1615 is still turned off since the diode 1615 disposed across the negative buck switch 1545 allows the current to flow from the source 1580 into the inductor 1500. Given that the current flow is now being carried by the buck switch in a positive direction 1505, the positive freewheel switch 1520 is then disabled 1635. At this point, the negative buck switch 1615 is engaged 1640.

[0016] When transitioning from the buck state to the freewheel state in the case where positive current 1505 is flowing into the buck inductor 1500 from the source 1580, the buck switch and freewheel switch elements are controlled in a symmetrical manner relative to the transition to the buck state from the freewheel state. As such, current flowing 1505 from the buck switch must be maintained in order to prevent discontinuous current flow through the buck inductor 1500. Accordingly, the negative buck switch 1545 is turned off 1645. Even though the negative buck switch 1545 is turned off, the diode 1615 disposed across the negative buck switch 1545 continues to allow current to flow from the source 1580 into the buck inductor 1500. The positive freewheel switch 1520 is then turned on 1650. This now enables positively flowing current 1510 to be directed into the buck inductor 1500. However the freewheel current 1510 is not allowed back to the source 1580 because the negative buck switch 1545 is turned off and the current flow is prevented by the diode 1615 that is disposed across the negative buck switch 1545. At this point, the positive buck switch 1540 is turned off 1660. At this point, current is being carried by the diode 1610 disposed across the negative freewheel switch 1525 and by the positive freewheel switch 1520 by virtue of the fact that the positive freewheel switch 1520 has been turned on. Accordingly, the negative freewheel switch 1525 is then turned on 1655, thus completing the transition from the buck state to the freewheel state when current is flowing into the inductor 1500.

[0017] Figs. 1C and ID are, respectively, a pictorial diagram and a timing diagram that depicts the operation of one illustrative example method and apparatus when current is flowing in a negative direction through a buck inductor. In this illustrative example method, study of the timing diagram (Fig. ID) depicts the sequence for controlling the positive buck switch 1540, the negative buck switch 1545, the positive freewheel switch 1520 and the negative freewheel switch 1525. For the sake of comprehension, it is best to examine the state of current flow during a freewheeling state depicted in the figure at point 1700. At this point, both of the freewheel switches are enabled and both of the buck switches are disabled. According to this illustrative use case, freewheel current 1513 is flowing out of the inductor 1500 and down through the synchronous freewheel switches comprising the negative freewheel switch 1525 and the positive freewheel switch 1520.

[0018] In this state, transitioning from the freewheel state to the buck state comprises a first step of disabling 1705 the positive freewheel switch 1540. By disabling the positive freewheel switch 1520, the freewheeling current 1513 is maintained because the negative freewheel switch 1525 is still turned on and current flow is maintained by the positive freewheel diode 1611 disposed across the positive freewheel switch 1520. At this point, the negative buck switch 1545 is enabled 1710. This now allows current 1507 to start flowing from the buck inductor 1500 back to the source 1580. It should be appreciated that turning on the negative buck switch 1545 allows the current to bypass the diode 1615 disposed across the negative buck switch and continue to be directed to the source 1580 by means of the diode 1620 disposed across the positive buck switch 1540. Once this current path is established, then the negative freewheel switch 1525 is disabled 1715. In an additional step, once both the positive and negative freewheel switches are disabled, the positive buck switch 1540 is then enabled 1720.

[0019] When transitioning from the buck state back to the freewheel state, the switches comprising the buck switch to the synchronous freewheel switch are controlled in a manner symmetrical to the manner in which the switches are controlled when passing from the freewheel state to the buck state. Accordingly, as negative current 1507 is flowing from the buck inductor 1500 back to the source 1580, transitioning to the freewheel state comprises a first step of disabling 1725 the positive buck switch 1540. This allows negative current 1507 to continue back to the source 1580 until the negative freewheel switch 1525 is enabled 1730. At this point the negative freewheel switch 1525 allows current 1513 flowing from the inductor 1500 to pass through the freewheel switch and as should be appreciated to current 1530 is also carried by the diode 1611 disposed across the positive freewheel switch 1520. At this point, the negative buck switch 1545 is disabled 1735. Once both buck switches are off, the positive freewheel switch 1520 is then enabled 1740. [0020] In operation, the controller 1570 determines the direction of current flow by means of the current sensor 1560 disposed so as to enable determination of the direction of current flow in the buck inductor 1500. It should be appreciated that the controller 1570 embodies the methods described herein for controlling the buck switches and the freewheel switches in a manner as described herein based upon the direction of current flowing in the buck inductor 1500. It should be appreciated that the sequence described is best followed as rapidly in succession as allowed by the turn on and turn off delays associated with the positive and negative buck switches and positive and negative freewheel switches. Furthermore, in one illustrative alternative method and apparatus, determination of direction of current flow is accomplished as soon as practical relative to the transition from a buck state to a freewheel state and relative to the transition from a freewheel state to the buck state.

[0021] It should also be appreciated that the direction of current flow may change from positive to negative or negative to positive during a buck state or during a freewheel state. Accordingly, even though a particular sequence for controlling the individual buck switches and individual freewheel switches is utilized when entering either a buck state or a freewheel state, an alternative sequence for controlling the individual buck switches and individual freewheel switches is utilized in the event that the direction of current flow changes during the interval of time within a particular buck state or a particular freewheel state. In lay terms, even though the transition from a freewheel state to a buck state follows the sequence for positive current flow, the transition to the alternative state, according to this alternative example method and apparatus, will follow the sequence for transitioning based on negative current flow when such a reversal of current flow is detected during a particular buck state or during a particular freewheel state. [0022] Fig. 2 is a pictorial diagram that depicts the operation of four control signals controlling four switches in an AC buck converter. Again referring to Fig. 1, there are generally four control signals that control the operation of one of two different buck switches 310 and 315 and two different synchronous freewheel switches 370 and 375. In normal operation, the positive buck signal 20 is substantially enabled to enable positive current to an inductor 360. By definition, a buck cycle is that portion of a pulse width modulation cycle where the positive buck signal 20 is active. Notice that in Fig. 2 a rising slope of the current waveform 30 is distinctly coincident with the rising edge of the buck signal 20. It is more correct to appreciate that the rising edge of the buck signal 20 enables the positive buck switch 310 and this then enables current to the inductor 360.

This in turn causes the slope of the current wave 30 to transition from a falling current level to a rising current level. It should also be appreciated that the timing of the positive buck signal 20 and the sequence with respect to the other signals is different in the negative current case 90 relative to the positive current case 80. In that case, the negative buck signal indicates that a buck cycle is in progress.

[0023] At the beginning of a buck cycle, i.e. when the positive buck signal is active, there is a window of vulnerability when the negative freewheel signal 50 is disabled and the window of vulnerability continues until the positive buck signal 20 is enabled. There is likewise a window of vulnerability at the end of the buck cycle after the positive buck signal 20 is disabled and the negative freewheel signal is again re-enabled. The reason for the vulnerability with is that in the positive current case 80 the switching sequence is intended to ensure continuity of conduction in the inductor 360. However when crossing the zero-current line the current wave 30 may actually change polarity during the window of vulnerability. In this case there is a discontinuity of current flow in the inductor. [0024] By example, which is symmetrical for all illustrative use cases, after the positive buck signal 20 has been disabled at the end of a buck cycle current should be flowing from the common terminal 340 toward the inductor 360 and out to the output terminal 365. Noticing the sequence of timing of the control signals, positive current flow 377 is supported by the intrinsic body diode 371 of the negative freewheel switch 370. Positive freewheel current 377 is further supported by the fact that the positive freewheel signal 40 has been enabled in order to turn on the positive freewheel switch 375. However, if during the window of vulnerability beginning with the termination of the positive buck signal 20 and continuing to enablement of the negative freewheel signal 50 a reversal of current flow occurs in the inductor 360, a negative current flow 378 cannot be supported because the negative freewheel switch 370 has not yet been turned on and the negative current flow 378 cannot be supported by the back biased body diode of the negative freewheel switch 370. Accordingly a discontinuity of current flow in the inductor 360 occurs. This leads to catastrophic failure of semiconductor switches as the back electromotive force in the inductor 360 results in high- voltage transients because there's no continuous current path for the current in the inductor 360.

[0025] A similar window of vulnerability exists at the end of a freewheel cycle. A freewheel cycle is the period of time where the negative freewheel signal 50 is active. Again referring to a positive current case 80 positive current flow 377 is supported by the fact that even though negative freewheel switch 50 is disabled positive current flow supported by the body diode 371 included in the negative freewheel switch 370. Since the positive freewheel signal 40 remains asserted beyond the assertion of positive buck signal 20 continuous current flow is supported in the inductor 360. However if there is a reversal of current flow in this window of vulnerability catastrophic results occur as previously described.

[0026] Figs. 3 and 4 are pictorial diagrams that depict simulation results

demonstrating loss of commutation. In Fig. 3, reversal of current flow occurs at point 100. In Fig. 3, current is negative and is diminishing (toward zero from a negative level) during a freewheel cycle. At some point during the window of vulnerability as described heretofore the current crosses the zero-current line 105 and results in discontinuity in current flow. Fig. 4 shows loss of commutation where discontinuity in current flow in an inductor occurs at the end of a freewheel cycle at point 120 and at the end of a buck cycle point 130. At the end of the freewheel cycle, current is positive and diminishing (from a positive level) when it crosses the zero-line at point 120. Prior to that point, the negative freewheel signal is turned off and the positive buck signal has not yet been enabled.

Current discontinuity in the inductor 360 occurs. At the end of a buck cycle, current is negative and is again increasing toward the zero-line when the zero-line is reached at point 130. Here, the positive buck signal has been disabled just prior to crossing the zero- line and the negative freewheel signal has not yet been enabled. Again, current discontinuity in the inductor 360 occurs.

[0027] Fig. 5A is a timing diagram that further clarifies current reversals during the termination of a freewheel cycle. As already described, the sequence of controlling the negative freewheel switch 50, the positive buck switch 20, the positive freewheel switch

40 and the negative buck switch 60 depends upon the direction of current flow in the inductor 360. In Fig. 5A, the current flowing in the inductor 360 is positive during a freewheel cycle and is diminishing toward zero (depicted by segment 150). Just prior to the falling edge of the negative freewheel signal 50, a decision is made that the current is in fact positive and the positive current switch sequence is selected. Accordingly, the first transition is the falling edge of the negative freewheel signal 50. When this occurs, a window of vulnerability 33 begins. As the current (segment 150) continues to diminishing toward zero it may in fact cross the zero line 105 during the window of vulnerability 33. This results in a loss of commutation 155 because, as already described above, there is no means for negative current to be carried by the freewheel switches and the positive buck switch has not yet been enabled by the positive buck signal 20. In commutation is again restored 160 once the positive buck signal 20 is enabled.

[0028] Fig. 5B is a timing diagram that further clarifies current reversals during the termination of a buck cycle. In Fig. 5B the current is negative just prior to the termination of a buck cycle (depicted by segment 165). It should be appreciated that because the current is negative a different termination sequence is selected. In this alternative termination sequence, as already described above for the case of negative current, it is the negative buck signal 60 that is disabled resulting in the beginning of a window of vulnerability 33. Began, and as the current is negative and is diminishing toward zero it may in fact cross the zero line 105 during the window of vulnerability 33.

This again results in a loss of commutation 170. In commutation is restored 175 once the positive freewheel signal 40 is enabled. [0029] Fig. 6 shows one example method for preventing catastrophic failure by preventing current reversals during the termination of a buck cycle or the termination of a freewheel cycle based on sensing the level of current while in an extended state. Here, a zero-line 105 indicates the reversal in polarity of current flow of an inductor 360 included in an output filter of an AC buck converter. In this example method, a "low-current window" 140 is established. In one is a example method were a buck cycle 150 terminates within a low current window 140 the buck cycle 150 is extended 35 until the level of current flowing in the inductor is outside of the low current window 140. Such extension of the buck cycle according to one example embodiment is accomplished by a merely extending the buck cycle using the buck control signal 20 and adjusting the timing all of the other control signals i.e. negative freewheel 50 positive freewheel 40 negative buck 60 according to the teachings of the incorporated reference. And in yet another example method, where a freewheel cycle 160 terminates within the low current window 140 the freewheel cycle is extended until the current flowing in the inductor is outside of the low current window 140. Again, this can be accomplished by manipulating the negative freewheel signal so as to extend the freewheel cycle until the current flowing in the inductor is outside of the low current window 140.

[0030] Fig. 7 shows one alternative example method for preventing current reversals during the termination of a buck cycle or the termination of a freewheel cycle by predicting when current flow in an output inductor is not within a low current window.

In this example method, a "low-current window" 140 is also established. In this alternative example method when a buck cycle 150 terminates within a low current window 140 the buck cycle 150 is also extended 35. However, in this alternative method, the buck cycle continues to be extended for an amount of time consistent with a projection as to how long the current in the output inductor 360 will require to be at a level outside of the low current window. Such extension of the buck cycle according to one example embodiment is accomplished by a merely extending the buck cycle using the buck control signal 20 and adjusting the timing all of the other control signals (i.e.

negative freewheel 50 positive freewheel 40 negative buck 60) according to the teachings of the incorporated reference. And in yet another example method, where a freewheel cycle 160 terminates within the low current window 140 the freewheel cycle is extended for an amount of time until the current flowing in the inductor is predicted to be outside of the low current window 140. Again, this can be accomplished by manipulating the negative freewheel signal so as to extend the freewheel.

[0031] Fig. 8 is a flow diagram that depicts one example method for ensuring continuous current flow in an output inductor included in an AC buck converter.

According to this example method, ensuring continuous current flow in the output inductor is accomplished by a method that comprises establishing a low current window relative to a zero line representing the polarity of current flow in the output inductor (step

200). Accordingly, the level of current flow in the inductor is detected (step 205). In the event that level of current flowing in the output inductor is within the low current window (step 210), then a buck cycle is extended 215. The buck cycle is extended until the level of current is no longer within the low current window (step 220). At that point the buck cycle is terminated (step 225).

[0032] Fig. 9 is a flow diagram that depicts one alternative example method for ensuring continuous current flow in an output inductor. In this alternative example method, when the level of current flowing in the output inductor is within the low current window (step 230), then a freewheel cycle is extended (step 235). The freewheel cycle continues to be extended so long as the level of current flowing in the inductor is within the low current window (step 240). Once the level of current is outside of the low current window then the freewheel cycle is terminated (step 245).

[0033] Fig. 10 is a flow diagram that depicts a hybrid method for ensuring continuous current flow in an output inductor. In this alternative example method, a low current window is also established (step 250). Again, the low current window is relative to a zero line that represents the polarity of current flow in the output inductor. In this alternative example method, the level of current flow in the inductor is detected (step 255).

Continuing further, When the level of current is within the low current window (step 260), then a buck cycle is extended (265). A prediction is then made as to how much time is required so that the current flowing in the inductor is no longer inside of the low current window. When the current is predicted to be outside of the low current window (step 270), then the buck cycle is terminated (step 275). [0034] Fig. 11 is a block diagram that depicts one example embodiment of a controller that ensures continuous conduction in an output inductor. According to this example embodiment, a control circuit 335 (as depicted in Fig. 1) includes a pulse with modulation (PWM) controller 400. This example embodiment further includes a polarity detector 425 and a window comparator 415. In operation, the PWM controller 400 receives a duty factor 440 and generates, according to send a duty factor, for control signals including a negative freewheel signal 50, a positive buck signal 20, a positive freewheel signal 40, and a negative buck signal 60.

[0035] The PWM controller 400 accepts a polarity signal 435, which is generated by the polarity detector 425. The polarity detector determines polarity based on a current sense signal (I-sense) that is received from the current detector (1560 in Fig. 1A). The I- sense signal is also provided to the window comparator 415. The window comparator 415 provides an indication 430 when the value of the I-sense signal is within a positive and negative threshold wherein said thresholds straddle zero current.

[0036] In normal operation, the PWM controller 400 generates timing sequences for the four switch control signals in accordance with the teachings and provided above, namely the descriptions of the apparatus and signal timings depicted in Figs. 1 A through ID. In this example embodiment, there is further included in a clock gate 405. So long as the output of the window comparator 415 remains inactive 411 the PWM controller 400 continues to generate a the control signals for negative freewheel 50, positive buck 20, positive freewheel 40, and negative buck 60. It should be appreciated that the PWM controller 400 determines the proper sequencing of these switch control signals as the transitions from a buck state to a freewheel state and from a freewheel state to a buck state based on the polarity signal 435 received from the polarity detector 425. [0037] At some point just prior to making a transition from a buck state to a freewheel state or from a freewheel state to a buck state, the PWM controller 400 causes of the state of the low current indicator 430 to be sampled by a sampling device 418, which is included one alternative embodiment of the present apparatus. The clock 410 is then gated off by the clock gate 405. The PWM and controller 400 will then clear the sampling device once the low current indicator 430 is no longer active. This 3 enables the clock 410 to the PWM controller 400 and the sequencing of the negative freewheel signal 50, the positive buck signal 20, the positive freewheel signal 40 and the negative buck signal 60 is allowed to continue.

[0038] While the present method and apparatus has been described in terms of several alternative and exemplary embodiments, it is contemplated that alternatives,

modifications, permutations, and equivalents thereof will become apparent to those skilled in the art upon a reading of the specification and study of the drawings. It is therefore intended that the true spirit and scope of the claims appended hereto include all such alternatives, modifications, permutations, and equivalents.