In the past, the semiconductor industry utilized various structures and methods to form power factor correction circuits. The power factor was generally recognized as a measure of the difference between the voltage and current waveforms of an alternating current (AC) waveform. Differences between the current and voltage waveforms resulted in low efficiency utilization of the supplied AC power. Power factor correction circuits were utilized to more closely align the shape of the current and voltage waveforms in order to provide higher efficiency. Examples of power factor correction circuits are disclosed in United States patent No.

5,134, 355 issued to Roy Alan Hastings on July 28,1992 ; patent No. 5,359, 281 issued to Barrow et al on October 25, 1994; and patent No. 5,502, 370 issued to Hall et al on March 26,1996 all of which are hereby incorporated herein by reference.

FIG. 1 illustrates a simplified schematic of a portion of a prior art power factor correction circuit 100 that had some similar functionality to the above referenced patents. Circuit 100 received an AC input voltage on inputs 112 and 113 and operated a switching transistor 101 to generate a DC output voltage on outputs 117 and 118. An error amplifier 102 and multiplier 103 operated together to form an AC reference voltage on an output 104 of multiplier 103. A current shaping block and

a logic and control section used the AC reference voltage to control transistor 101. The control signals to transistor 101 were designed to modify the input current from inputs 112 and 113 in a manner that made the shape of the input current waveform closely match the shape of the input voltage waveform in order to provide a power factor that approached a unity value.

One problem with the previous power factor control circuits was the transient response time. In order to provide low distortion in the input current waveform, the control loop of circuit 100 had a very slow response time.

In order to prevent input noise from affecting the output voltage, the bandwidth of the control loop generally was about ten times less than the frequency of the rectified AC input voltage. Typically the bandwidth was limited to about ten to twelve Hz (10-12 Hz). Because of the low bandwidth of the control loop and the low bandwidth resulting from the compensation of amplifier 102, circuit 100 had a slow response to transient voltages on inputs 112 and 113 which often resulted in an over-voltage or under-voltage condition at outputs 117 and 118. The over- voltage condition could result in damage to the load connected to outputs 117 and 118, while the under-voltage condition could result in shutdown of the load. Such transients often occurred and were very common at start- up.

Accordingly, it is desirable to have a method of forming a power factor correction device that has a wider loop bandwidth, that has an error amplifier that quickly responds to transients, and that provides greater protection for loads connected to the power factor correction device.

Brief Description of the Drawings FIG. 1 illustrates a simplified schematic of a portion of a prior art power factor correction circuit; FIG. 2 schematically illustrates a block diagram of an embodiment of a power system in accordance with the present invention; FIG. 3 schematically illustrates a portion of an embodiment of a power factor correction device in accordance with the present invention; FIG. 4 schematically illustrates a portion of an embodiment of an error amplifier of the power factor correction device of FIG. 3 in accordance with the present invention; FIG. 5 schematically illustrates a portion of another embodiment of a power factor correction device in accordance with the present invention; FIG. 6 schematically illustrates a portion of an embodiment of an error amplifier of the power factor correction device of FIG. 5 in accordance with the present invention; and FIG. 7 schematically illustrates an enlarged plan view of a semiconductor device that includes the power factor correction device of FIG. 3 in accordance with the present invention.

For simplicity and clarity of illustration, elements in the figures are not necessarily to scale, and the same reference numbers in different figures denote the same elements. Additionally, descriptions and details of well known steps and elements are omitted for simplicity of the description.

Detailed Description of the Drawings FIG. 2 schematically illustrates a simplified block diagram of a power system 10 that includes a power factor control function. Power system 10 receives an AC voltage and current from an AC power source 11 such as a standard office or home AC line outlet. The AC power is received on inputs 12 and 13 of a power factor control module 16.

Module 16 typically provides a DC voltage and current on outputs 17 and 18. A power conversion module 19 receives the DC voltage and converts the DC voltage to an output voltage that is suitable for driving a load. Module 19 may be a variety of linear or non-linear power conversion modules such as a switching power supply controller, a pulse width modulated (PWM) controller, a buck converter, or a boost converter. Module 16 adjusts the power factor of the incoming AC power in order to provide high efficiency power conversion. System 10 may be a power system of various applications including a power supply section of a computer system, a personal computer, or other electronic system.

FIG. 3 schematically illustrates a portion of an embodiment of a power factor correction module 25 that functions similarly to module 16 that is illustrated in FIG. 2. Module 25 includes a power factor correction circuit or power factor correction device 26 that functions to adjust the shape of the waveform of the incoming input AC current to closely match the waveform of the input AC voltage. Module 25 receives an AC input signal on inputs 12 and 13, and provides a DC output voltage and current on outputs 17 and 18. A rectifier 21 of module 25 rectifies the AC input signal and provides a full wave rectified AC input voltage to a switching transistor 43 and to a voltage input 55 of device 26. An inductor 24 connected between rectifier 21 and transistor

43 functions as an energy storage element that assists filtering the rectified AC input voltage. Power factor correction device 26 operates transistor 43 to generate the DC output voltage on outputs 17 and 18, and to adjust the shape of the input current received at inputs 12 and 13. A feedback network 49 provides a feedback voltage that is representative of the value of the output voltage on outputs 17 and 18. A filter capacitor 22 typically is connected between outputs 17 and 18 to reduce noise and ripple in the output voltage. An input voltage divider 44 forms an AC sense signal that has the same shape as the AC input voltage but has a lower amplitude, thus, the AC sense signal is representative of the AC input voltage.

The AC sense signal is applied to a sense signal input 45 of device 26. A current sense input 50 receives a current sense signal that represents the value of the AC input current. The current sense signal is derived from the current through a current sense resistor 57 that is connected between output 18 and rectifier 21. Output 18 is also connected to a voltage return 54 of device 26.

Output 18 and voltage return 54 form a common potential for module 25 and often is used as a ground reference potential. Rectifier 21, feedback network 49, and divider 44 are illustrated by dashed boxes.

Power factor correction device 26 includes a sample- and-hold error voltage generator 36, a voltage reference circuit 27, an over-voltage amplifier 33, an under-voltage amplifier 34, an analog reference multiplier 46, a current shaping block 48, and a logic and output driver section 47. Sample-and-hold error voltage generator 36 includes an error amplifier 32, an analog switch 38, a storage element 39, an amplifier 41, and a zero crossing detector 37. Sample-and-hold error voltage generator 36 is formed to have a bandwidth that is no less than, and preferably is approximately equal to, the frequency of the AC sense

signal applied to input 45. This frequency is hereinafter referred to as the zero crossing frequency. Typically the zero crossing frequency has a value that is no less than and preferably is approximately equal to twice the frequency of the AC input signal applied to inputs 12 and 13. In other embodiments it may have other values.

In operation, error amplifier 32 is formed to receive both the first reference voltage from circuit 27 and the feedback voltage formed by feedback network 49, and to responsively generate a deviation voltage on an output of amplifier 32. The deviation voltage is representative of the deviation of the output voltage value from the desired output voltage value. Detector 37 is formed to detect each zero crossing of the AC sense signal applied to input 45 and responsively enable switch 38 to connect the output of amplifier 32 to storage element 39 in order to store or adjust the value of the deviation voltage on element 39 at each zero crossing of the AC sense signal. In the preferred embodiment, the AC sense signal is representative of the AC input signal, thus, switch 38 is enabled at each zero crossing of the AC input signal applied to inputs 12 and 13. Detector 37 typically enables switch 38 within at least ten to twenty degrees of each zero crossing. It is important to enable switch 38 substantially during the zero crossing because the output of multiplier 46 has a zero value during this time, thus, changing the value of node 40 during this time does not have a detrimental effect on the output voltage or the power control function. In the preferred embodiment, detector 37 detects each zero crossing and enables switch 38 within no greater than about three degrees of the zero crossing. Switch 38 can be a variety of well known electronic switches and preferably is a metal oxide semiconductor (MOS) coupler device. Such zero crossing detectors and switches are well known to those skilled in

the art. Amplifier 41 is formed to receive the stored value of the deviation voltage from storage element 39 and responsively form an error voltage on an output 42 of generator 36. Preferably, amplifier 41 is formed as a unity gain buffer that has a high input impedance in order to prevent disturbing the value of the deviation voltage stored on element 39. Amplifier 41 generally has an input current of less than ten nano-amps. In other embodiments the gain of amplifier 41 may be different.

Multiplier 46 receives the value of the error voltage from output 42 and multiplies the value of the AC sense signal by the value of the error voltage in order to form an AC reference signal on an output of multiplier 46. The AC reference voltage has the same shape as the waveform of the AC input voltage but has an amplitude that is adjusted by the value of the error voltage. Current shaping block 48 receives the AC reference voltage from multiplier 46 and receives the current sense signal from input 50 and responsively controls driver section 47. The functions of current shaping block 48, logic and output driver section 47, and transistor 43 are well known to those skilled in the art.

Because generator 36 is sampling the deviation voltage at a frequency of no less than twice the frequency of the AC input signal applied to inputs 12 and 13, the bandwidth of generator 36 is limited to this frequency, thus, amplifier 32 can have a much wider uncompensated bandwidth. The uncompensated bandwidth of amplifier 32 generally is between one and ten MHz (1-10 MHz) but may be wider. The wide bandwidth of amplifier 32 allows the bandwidth of generator 36 to be controlled by the zero crossing frequency of the signal applied to detector 37.

Additionally, the wide bandwidth of amplifier 32 ensures that the output of amplifier 32 quickly changes the deviation voltage as the output voltage varies thereby

providing a more accurate output voltage value. Because of the wide bandwidth of amplifier 32, the output signal of amplifier 32 may have some ripple. However, the value of the deviation voltage is only stored at the zero crossings consequently this ripple is removed. The wide bandwidth of amplifier 32 and of generator 36 provides a fast control loop response that can quickly respond to changes in the output voltage or current due to demands from the load. Device 26 typically provides a response to changes in the output voltage that is ten to twenty (10- 20) times faster than prior art power factor correction circuits. Those skilled in the art will note that amplifier 32 typically will have some external compensation that limits the closed loop bandwidth of amplifier 32 to something that is less than the uncompensated bandwidth but typically is still no less than the zero crossing frequency.

Over-voltage amplifier 33 receives both a second reference voltage from circuit 27 and the feedback voltage and generates an output signal when the value of the output voltage exceeds the desired upper limit value. The output of amplifier 33 is connected to storage node 40 and changes the value of the deviation voltage stored on element 39 when the value of the output voltage exceeds the desired upper limit. Similarly, under-voltage amplifier 34 receives both a third reference voltage from circuit 27 and the feedback voltage and responsively provides an output signal when the value of the output voltage is less than the desired lower limit. The output of amplifier 34 is also connected to node 40 and changes the value of the deviation voltage stored on element 39 when the value of the output voltage is less than the desired lower value. The new stored voltage resulting from over-voltage or under-voltage condition quickly changes the AC reference voltage and the corresponding

control voltage to transistor 43 to modify the value of the output voltage. The response to such transients is not limited by the bandwidth of amplifier 32 nor by the sampling rate of generator 36. It should be noted that at the next zero crossing, the value of the deviation voltage will be restored on element 39, thus, the over-voltage or under-voltage modification is very brief (generally no greater than one-half of a cycle). Forming the over- voltage and under-voltage circuits to change the value the stored value of the deviation voltage provides a method of quickly responding to the over or under voltage condition without being limited by the bandwidth of the error amplifier. In order to prevent the output of amplifiers 33 and 34 from affecting the voltage value stored on element 39 during periods of normal operation, the output drivers of amplifiers 33 and 34 are formed as a sink only and a source only stage, respectively. Thus, amplifiers 33 and 34 are formed to generate an output current only during an over or under voltage event, respectively.

Those skilled in the art will recognize that amplifiers 33 and 34 may also be formed as comparators.

FIG. 4 schematically illustrates a preferred embodiment of portions of amplifier 32 that is illustrated in FIG. 3. Amplifier 32 typically includes a differential amplifier 60 and an output stage. In this preferred embodiment, amplifier 32 is a transconductance amplifier and the output stage includes a current source 61 and a current sink 62 that have current output and input, respectively, values that are controlled by the differential outputs of amplifier 60. Using a transconductance amplifier facilitates quickly charging or discharging element 39 when switch 38 is enabled.

Referring back to FIG. 3, amplifier 32 has an inverting or first input connected to a first output 31 of circuit 27, a second input or non-inverting input

connected to feedback network 49, and an output connected to a first terminal of switch 38. Detector 37 has an input connected to input 45 to receive the AC sense signal, and an output connected to a control electrode of switch 38. A second terminal of switch 38 is connected to storage node 40. Storage element 39 has a first terminal connected to node 40 and a second terminal connected to voltage return 54. Amplifier 41 has a non-inverting terminal or first terminal connected to node 40, and an output connected to a second input terminal of amplifier 41. Over-voltage amplifier 33 has a non-inverting input or first input connected to a second output 28 of circuit 27, an inverting input or second input connected to feedback network 49, and an output connected to node 40.

Under-voltage amplifier 34 has a non-inverting input or first input connected to a third output 29 of circuit 27, an inverting input or second input connected to feedback network 49, and an output connected to node 40.

Multiplier 46 has a first input connected to output 42 of generator 36, a second input connected to input 45, and an output connected to an input of block 48.

FIG. 5 schematically illustrates a portion of an embodiment of a power factor correction module 65 that is an alternate embodiment of module 25 that is illustrated in FIG. 3. Module 65 includes a power correction device 75 and an error voltage generator 76 that respectively replace device 26 and generator 36. Generator 76 utilizes an error amplifier 66 that has a selectable output stage as will be seen in more detail hereinafter.

FIG. 6 schematically illustrates a preferred embodiment of portions of amplifier 66 that is illustrated in FIG. 5. For clarity, this description will have references to both FIG. 5 and FIG. 6. Amplifier 66 is formed to receive both the first reference voltage from circuit 27 and the feedback voltage from network 49, and

to responsively generate the deviation voltage on an output of amplifier 66. Amplifier 66 preferably is formed as a transconductance amplifier that has a differential amplifier stage and an output stage having a plurality of sink only current sources and a plurality of source only current sources. A variable source only current source 67 and a variable sink only current source 68 are controlled by the differential outputs of amplifier 60 similarly to sources 61 and 62 illustrated in FIG. 4. Amplifier 66 also has a plurality of selectable current sources including a selectable source only current source 69 and a selectable sink only current source 70 that are selectively switched or enabled in parallel with sources 67 and 68, respectively, by detector 37 during the zero crossing of the AC input signal. Switches 71 and 72 are enabled by detector to implement the switchable parallel connection. Switches 72 and 71 are similar to switches 38 shown in FIG. 3. The current capacity of sources 69 and 70 are controlled by the differential outputs of amplifier 60. During the zero crossing, detector 37 enables switches 71 and 72, and sources 67,68, 69, and 70 adjust the value of the error voltage on element 39 in response to the inputs to amplifier 66. Sources 67 and 68 have smaller current values than sources 69 and 70 so that when switches 71 and 72 are not enabled sources 67 and 68 are controlled by amplifier 60 to change the value of the deviation voltage stored on element 39. Sources 69 and 70 generally have a current capacity of at least twice the capacity of source 67 and 68 and preferably at least ten times the capacity. Forming amplifier 66 to have a selectable output stage maintains the deviation voltage value on the output of amplifier 66 and minimizes the voltage range over which the output must slew when detector 37 enables amplifier 66. Since the output of amplifier 66 is always connected in the control loop of

module 65 the feedback is maintained thereby maintaining the control provided by module 65.

In order to facilitate this operation, amplifier 66 has an inverting or first input connected to a first output 31 of circuit 27, a second input or non-inverting input connected to feedback network 49, an output connected to node 40, and a switch control input connected to the output of detector 37.

FIG. 7 schematically illustrates an enlarged plan view of a portion of an embodiment of a semiconductor device 80 that is formed on a semiconductor die 81. Power factor correction device 26 is formed on die 81.

In view of all of the above, it is evident that a novel device and method is disclosed. Forming the power factor correction device to sample the output of the error amplifier at a frequency no less than twice the input frequency allows the error amplifier to have a wide bandwidth while limiting the control loop bandwidth to a frequency no greater than twice the input frequency.

Coupling the over-voltage and under-voltage comparators to adjust the stored value from the error amplifier when the over or under-voltage condition occurs results a response to such conditions that is at least ten times fast than the response provided by prior power factor control circuits.