Modular structure of an apparatus for regulating the harmonics of current drawn from power lines by an electronic load

[0002] This application is a continuation of application Ser. No. PCT/IL99/00380, filed Jul. 12, 1999.

[0003] The present invention relates to the field of power electronics. More particularly, the invention relates to a method for regulating the harmonics of current drawn from power lines, by electric equipment, using modular design and construction techniques.

[0004] The art has devoted a considerable effort to the problem of shaping the current amplitude and phase drawn from the power line, loaded by electric equipment. Switch Mode and Resonant Converters are widely used for DC—DC (Direct Current-Direct Current), DC-AC (Direct Current-Alternate Current), AC-DC and AC—AC conversion. In some applications, the purpose of power conversion is to shape the current at the input of the converter. For example, an input power stage known in the art as an Active Power Factor Correction (APFC) circuit, the function of the converter is to insure that the AC current supplied from the source (power line) will appear in almost the same shape and phase as the source voltage. Thus, a Power Factor (PF) of unity is obtained.

[0005] The need for APFC stages is driven by the worldwide concern about the quality of power supplies. Generally, poor power factors results is several problems, comprising low efficiency of power transmission, interference to normal operation of other units connected to the same power line, as well as distortion of the voltage shape. Voluntary and mandatory standards that restrict the acceptable level of the current harmonics generated by the loading equipment, have been adopted by many countries, so as to maintain high power quality.

[0006] Another advantage of APFC is increased power levels that may be drawn from the power line. Looking at the spectral components of the current without APFC, only the fundamental (first) harmonic component contributes to the real load power. The root mean square (rms) magnitude of the current is increased beyond the level of the fundamental magnitude, as a result of higher order harmonics. Since the operation of protection elements, such as fuses and circuit breakers, is affected by the rms current, increased rms level limits the maximum power level that may be drawn from the line. Using electric equipment with APFC enables the rms current level to be equal to the level of the fundamental harmonic. Hence, the power that may be drawn from the line may reach its maximum theoretical level.

[0007] From the above description, it is clear that the need for APFC is wide spread, and economic implementation of APFC circuitry is of prime importance. Moreover, the cost of APFC circuits is of great concern considering the fact that APFC cost is an add-on expense to the functionality of the original equipment. As a result, APFC units should be constructed using methods that are economic during production, and are easy to integrate in any existing equipment.

[0008] Some known methods for physical implementation of APFC stages, sub-circuits, which are part of the electrical circuit of the (loading) equipment are used. In this case, the designer of this equipment engages passive and active components, as well as Integrated Circuits (ICs) to construct the desired APFC stage. Normally, designs that involve many components suffer from several drawbacks, like relatively high cost and low reliability. Moreover, these designs require many wiring connections, resulting in relatively high susceptibility to Electro-Magnetic Interference (EMI), “ground noise” and other engineering disadvantages.

[0009] Other construction methods comprise one “block” (APFC block) that includes all the components and circuitry of an APFC stage. This implementation provides functional solution to the design problem from the electrical aspect, but still requires satisfactory heat dissipation design. Still, a separate APFC unit should be normally purchased from another manufacturer (which is more skilled in this kind of products), leading to higher cost. Another disadvantage is the fact that this realization is not compatible with IC technology, which normally has the benefit of low cost in mass production. Further problem using APFC block is the fact that in this implementation, all heat dissipating components, such as the main switch, diodes and the inductor are close to each other. This proximity leads to a severe heat dissipation problem, which may limit the power handling capability of such module. Additionally, because of this component proximity, EMI restrictions require heavy snubbing, shielding and filtering, all leading to complexity, higher cost and lower efficiency.

[0010] U.S. Pat. No. 5,530,635 to Yashiro describes a power supply divided into a noise filtering module, one or more power factor and harmonic correction modules, one or more DC—DC converter modules and a backup power supply for the power converter modules. Each module may optionally be combined so as to construct a desirable power supply according to various required specifications. However, this construction lacks simplicity, since it is not realizable using IC technology and in addition is not advantageous with respect to optimal heat dissipation.

[0011] U.S. Pat. No. 5,047,912 to Pelly describes a modular four terminal realization of APFC stages. The control circuit comprises a signal differentiator, for generation of a reference signal to the feedback loop of the circuit. Differentiators are known in their extreme noise sensitivity, which may corrupt the output signal. Switching circuits such as Pulse Width Modulation (PWM) Boost converters, which are characterized by high frequency noise injection, are also problematic to use. Another problem arises from this patent is the reference feedback signal is driven from the line voltage, which contains additive noise that distorts the shape of the desired controlled line current.

[0012] A different problem, which is not solved by U.S. Pat. No. 5,047,912 is the generation of a local power supply (from the line voltage) for the control circuitry. High power APFC stages requires local power supplies of several watts. Using an external power supply for this purpose requires additional circuitry and pin connection, which increase cost. Another approach is to drive the power supply from a voltage divider from of the output voltage, or from the voltage across the main switch. This approach is extremely inefficient, especially in case of high power applications.

[0013] Another disadvantage of U.S. Pat. No. 5,047,912 is the need for fast diodes for its input rectifier, due to the fact that the line rectifier in this patent is locate after the input inductor. At this point, the signals are switching frequency signals that requires fast rectifier diodes, so as to eliminate high power loss. Using fast diodes introduces higher switching noise, as well as higher cost (since such diodes are more expensive).

[0014] All the methods described above have not yet provided satisfactory solutions to the task of constructing simple, low-cost, modular APFC stages.

[0015] It is an object of the invention to provide an APFC unit, for regulating the harmonics of power line current, that comprise low number of components.

[0016] It is another object of the invention to provide a compact, low cost APFC unit.

[0017] It is another object of the invention to provide a compact APFC unit, with a minimal amount of interconnections.

[0018] It is still another object of the invention to provide an APFC unit with efficient local power supply feeding switching and control circuit.

[0019] It is yet another object of the invention to provide an APFC unit which is compatible with IC technology.

[0020] It is yet another object of the invention to provide an APFC unit which do not comprise signal differentiation, while overcome the drawbacks of the prior art.

[0021] Other purposes and advantages of the invention will appear as the description proceeds.

[0022] The invention is directed to construction of a modular apparatus for regulating the harmonics of current drawn from power lines by electric equipment. This apparatus comprises a first rectifier circuit module consisting of an array of rectifying diodes, the inlet of which is connected to the power lines; a second Switch and Controller Assembly (SCA) module for controlling said current to follow the voltage across said power lines by sampling the current drawn from said power lines and the output voltage of said second SCA. The second SCA module consists of a power switch and controller and their interconnections, the inlet of which is connected to the outlet of said first rectifier circuit module through at least one serially connected input inductor, for filtering the current drawn from the rectifier module; and at least one output capacitor connected in parallel with the output of said SCA and said load, for filtering the output voltage ripple at said load.

[0023] Preferably, the main power factor correction operation is carried out by the SCA module. Correction is obtained by a high frequency controlled switching of the input rectified voltage. Further rectification is carried out by a high frequency rectifier diode connected in series with the SCA outlet. The control circuit samples both the input current and the output voltage of the SCA and controls the switching transistor accordingly by an inherent pulse width modulator responsive to the control signal. The control circuit is fed by a power supply via an inlet in the SCA module. Preferably, optional soft switching is obtained using an integrated turn-on snubber, consists from an inductor, a capacitor and two rectifier diodes, connected between the input and output of the SCA. These two diodes are integrated into the SCA module.

[0024] Preferably, the SCA module contains only semiconductor components, so as to be fabricated as a monolithic integrated circuit for low power applications, or as a hybrid circuit for medium and high power applications, in both cases packaged in a three terminal or two port heat dissipating package. Using this construction, increased power handling capability is obtained. This construction also reduces the number of interconnections required for integration, as well as reducing the unit cost of mass production. Passive components required for SCA operation, like inductors and some of the capacitors, which can not be fabricated as semiconductor products are connected via suitable inlets in the SCA package in case of monolithic design, or within the module in case of hybrid design.

[0025] Preferably, the proposed construction enables separate mounting of modules and components in a manner that is optimal to heat dissipation. Interconnections between different terminals of models or components of this construction are provided using wires and/or conductive traces on a printed circuit board or on an alumina substrate, reducing the labor associated with low cost construction. The apparatus further comprises an internal power supply driven from the output voltage or from the switched voltage of the SCA module.

[0026] The above and other characteristics and advantages of the invention will be better understood through the following illustrative and none-imitative detailed description of preferred embodiments thereof, with reference to the appended drawings, wherein:

[0027] FIG. 1 is a schematic diagram of a conventional (prior art) PWM boost converter;

[0028] FIG. 2 is a schematic diagram of a conventional (prior art) APFC stage;

[0029] FIG. 3 is a schematic diagram of a conventional (prior art) APFC stage in one block;

[0030] FIG. 4 is a schematic diagram (prior art) of a control circuit of an APFC stage;

[0031] FIG. 5 is a schematic diagram (prior art) of another control circuit of an APFC stage;

[0032] FIG. 6 is a schematic diagram of a modular construction of an APFC stage, according to the present invention;

[0033] FIG. 7 is a schematic diagram of a modular construction of an APFC stage, as a two-port network;

[0034] FIG. 8 is a schematic diagram of the construction illustrated in FIG. 6, with improved heat dissipation;

[0035] FIG. 9 is a schematic diagram of a Switch and Control Assembly (SCA), according to the present invention;

[0036] FIG. 10 is a schematic diagram of a resonant local power supply for the SCA, according to the present invention;

[0037] FIG. 11 is a schematic diagram of a resonant local power supply integrated into the SCA, according to the present invention;

[0038] FIG. 12 is a schematic diagram of an SCA with a lossless snubbing circuitry, according to the present invention;

[0039] FIG. 13 illustrates the APFC stage of FIG. 6, constructed according to the present invention, where the SCA is fabricated as a four pin IC;

[0040] FIG. 14 illustrates the APFC stage of FIG. 7, constructed according to the present invention, where the SCA is fabricated as a five pin IC;

[0041] FIG. 15 illustrates an APFC stage, constructed according to the present invention, where the SCA is fabricated using hybrid technology;

[0042] FIG. 16 illustrates an APFC stage, constructed according to the present invention, where the SCA is constructed in a manner suitable for hybrid IC technology;

[0043] FIG. 17 illustrates a full APFC stage construction according to the present invention.

[0044] For a better understanding of the present invention, the structures of prior art are examined, according to the illustrations presented by FIGS. 1 to 5. A typical prior art implementation of APFC stage, known as PWM boost converter, is shown in FIG. 1. The input AC voltage is rectified by an array 1, of four rectifying diodes, and fed into a boost stage that comprises an input inductor 2 (L_in), a switch 3 (S₁), a high frequency rectifier 4 (D₂), an output filtering capacitor 5 (C₀) and a load, represented by a resistor 6 (R_L). Switch 3 is driven by a high frequency control signal with duty cycle D, such that the shape of the input current i_infollows the shape of the rectified input voltage V_ivR. Consequently, the input terminal (before the inductor 2) will have resistive characteristics, i.e., the Power Factor (PF) equals unity.

[0045] Another prior art implementation, shown in FIG. 2, comprises both passive and active components. The input AC voltage is rectified by an array 7, of four rectifying diodes, and fed into a boost stage that comprises an input inductor 8 (L_in), a power Metal Oxide Semiconductor Field Effect Transistor (MOSFET) switch 9 (Q₁), a high frequency rectifier 10 (D₂), an output filtering capacitor 11 (C₀), a load, represented by a resistor 16 (R_L), an IC Active Power Factor (APF) controller 12 with its auxiliary passive components circuit 17, a current sensing resistor 13 (R_s), a voltage sensing resistor 18 (R_ac),and an output voltage divider, constructed from resistors 14 (R₁) and 15 (R₂). MOSFET switch 9 is driven by a high frequency control signal from the APF controller 12. Here, some components (such as inductor 8 and capacitor 11) are of relatively large size, which is not compatible with microelectronics technology. Other silicon semiconductors, such as APF controller 12 and rectifier 10 may be fabricated on a monolithic IC chip. However, the stage of FIG. 2 still comprises many discrete components with their associated wiring, that preclude the streamlining of the APFC stage. For example, the rectified input voltage V_ivRshould be sensed, and hence a wire should be connected between the feed point “c” and the controller 12. In order to regulate the output voltage, the voltage at point “b” (the output of the voltage divider R₁, R₂) is also sensed by the controller 12.

[0046] Another prior art, comprising a complete stage, similar to the stage of FIG. 2, (excluding the output filtering capacitor 20) packaged into one APFC block 19 is shown in FIG. 3. Here, all heat dissipating elements are in close proximity, complicating heat removal.

[0047] In an improved prior art APFC stage, illustrated in FIG. 4, comprises a modified control operation, in which sensing of the input voltage V_ivRis not required. The voltage V_aat point “a” is a pulsating voltage of maximum amplitude V₀and duration of T_OFFwhen switch 21 (S₁) is not conducting. Consequently, the average voltage at point “a” v_avis given by: $\begin{array}{cc}{\upsilon }_{\mathrm{a\upsilon }}=\frac{V_0·T_{\mathrm{OFF}}}{T_s}& \text{[Eq. 1]}\end{array}$

[0048] where T_sis the PWM switching period, or:

[0049] v_av=V₀·D_OFF  [Eq.2]

[0050] where $\begin{array}{cc}{D}_{\mathrm{OFF}}=\frac{T_{\mathrm{OFF}}}{T_S}& \text{[Eq. 3]}\end{array}$

[0051] Similarly, the “on” duty cycle D_ONwhile S₁is conducting (a period of T_ON) is given by: $\begin{array}{cc}{D}_{\mathrm{ON}}=\frac{T_{\mathrm{ON}}}{T_S}& \text{[Eq. 4]}\end{array}$

[0052] The input voltage V_ivRfed into the boost converter is assumed to be of low frequency, compared to the switching frequency f_s(f_s=1/T_s), and hence may be considered constant during one or several switching periods. If the power stage is properly controlled, the average low frequency voltage across inductor 22 (L_in) is almost zero (otherwise the current would risen to large values). Thus:

[0053] v_ivR=v_av  [Eq. 5]

[0054] where v_ivRis the instantaneous low frequency component of V_ivR.

[0055] From Eq. 1 and Eq. 5, the following equality is obtained:

[0056] v_ivR=V₀·D_OFF  [Eq. 6]

[0057] In case when D_OFFis programmed according to the following rule:

[0058] D_OFF=K·i_ina  [Eq. 7]

[0059] where K is a constant, and i_inais the low frequency component of the input current i_in, then:

[0060] v_ivR=V₀·K·i_ina  [Eq. 8]

[0061] or: $\begin{array}{cc}{i}_{\mathrm{ina}}=\frac{{\upsilon }_{i\upsilon R}}{V_0·K}& \text{[Eq. 9]}\end{array}$

[0062] If the filtering capacitor 23 (C₀) is sufficiently large so that the ripple of the output voltage V₀may be neglected, then according to Eq. 9 the input current follows the input voltage. Hence, the converter stage looks resistive with the following input resistance:

[0063] Re=K·V₀  [Eq. 10]

[0064] From Eq. 10, the value if the input resistance R_eas well as the input current may be controlled by adjusting the value of K. Practically, the output voltage should remain constant even for variations in the load 24 (R_L). This may be obtained by establishing a closed control loop on K. The voltage v_iin, which is proportional to the input current (i_ina) sensed by the sensing resistor 25 (R_s), is fed into a multiplier 26 (M). An error amplifier 27, compares the output voltage V₀with a reference voltage V_ref, generating an error voltage V_e, which is also fed into the multiplier 26. The product v_iin·V_e, which is proportional to i_inaby the factor K, is fed into a PWM modulator 28, which generates the duty cycle pulse (D_OFF) according to Eq. 7.

[0065] Another improved prior art APFC stage, illustrated in FIG. 5, exploits the current flowing in the MOSFET switch 29 to generate the duty cycle pulse (D_OFF). The sensing resistor 30 (R_s) is placed in series with the MOSFET switch 29, whereas the voltage across it is being used to measure the input current i_ina. This implementation, as well as the implementation described in FIG. 4 above, do not require input voltage sensing, lowering the number of the required interconnections.

[0066] According to one preferred embodiment of the present invention, the APFC stage comprises four basic sub-circuits, two modules and two structural elements, being connected to each other as illustrated in FIG. 6. The first module consists from a widely used input rectifier array 31 (D₁). This module has one inlet connected to the power line and one DC outlet. This inlet is normally pre-filtered by a conventional noise filter to suppress noise entering from the line. A capacitor (C₁) may be optionally connected across the rectifier array output so as to filter high frequency components of the rectified voltage to ground. The first structural element is the input inductor 32 (L_in), which may tapped and may consist from additional inductor, in case when a snubber is used. This inductor is being connected between the rectifier module and the SCA module.

[0067] The second module is the Switch and Control Assembly (SCA) 33, consists from all the semiconductor components of the switching, control, rectifier and power supply circuitry. The inlet of this module is being connected to the output of the first module 31 via the input inductor 32 (L_in), and the outlet to the load. The SCA module may optionally comprise a turn-on snubber. According to the invention, the SCA module may have three terminals, one for input, one for output and one for ground.

[0068] The second structural element is the output capacitor 34 (C₀), being connected across the outlet of the SCA, in parallel with the load. This modular construction is simpler than prior art implementations, with very few interconnections and with a simple forward flow of power and signals. In addition, the physical location of the three power dissipating elements, the input rectifier array 31 (D₁), the input inductor 32 (L_in), and the SCA 33 is not critical, so they may be mounted separately, with the best possible heat removal.

[0069] FIG. 7 illustrates a different construction, according to another preferred embodiment of the present invention. In this implementation, SCA 35 is connected to the same construction described if FIG. 6 above, but as a two port network. This construction provides sensing of the input current within the SCA, thereby simplifying the control circuitry.

[0070] According to another preferred embodiment of the present invention, the high frequency output rectifier diode 36 (D₂) is separated from SCA 37, as shown in FIG. 8. This construction allows better cooling and mounting.

[0071] The SCA may by realized differently with respect to its internal circuitry, packaging and production technology. FIG. 9 illustrates a typical realization of an SCA circuitry, according to the invention, where all microelectronics compatible components are grouped together. The internal power supply (V_CC) required for SCA operation may be obtained from the output voltage using an auxiliary circuit, which comprises a current source 38 (I_aux), a zener diode 39 (Dz_aux) for power supply stabilization, and a smoothing capacitor 40 (C_aux). In case of high power applications, the smoothing capacitor 40 (C_aux) may be an electrolytic capacitor located outside the SCA module. This SCA implementation employs a typical PWM modulator, controlled by a control circuit as described above with reference to FIG. 4.

[0072] According to the invention, a different internal power supply for high power modules is constructed, as shown in FIG. 10. A charge pump capacitor (C_x) is connected in series with the positive terminal of the pulsating voltage, which may be the drain of the switching transistor. A rectifying diode (D_x) is connected in series with the output of the charge pump capacitor. A zener diode (D_Zaux) for power supply output voltage stabilization is connected between the rectifier diode (D_x) and ground. A smoothing capacitor (C_aux) is connected in parallel with the zener diode (D_Zaux). An inductor (L_x) is connected in series with another rectifier diode (D_x1), between the second terminal of the charge pump capacitor (C_x) and ground.

[0073] According to one preferred embodiment of the invention, power supply output voltage is stabilized by a fixed or adjustable voltage regulator, instead of using the zener diode D_Zaux. This voltage regulator provides better ripple rejection at the output. The voltage regulator input is connected to the cathode of the rectifying diode D_x, the common terminal is connected to ground and the power supply output voltage is taken from the voltage regulator output, filtered by the smoothing capacitor C_aux.

[0074] The input of the internal power supply circuit may be connected to a pulsating source (V_pulse), such as the voltage V_adefined in FIG. 4 above. For each rising edge of the square-wave input, the charge pump capacitor 41 (C_x) delivers a constant charge, Q_x1, to the output V_CC, which is given by:

[0075] Q_x1=V_p·C_x  [Eq. 11]

[0076] where V_pis the magnitude of V_pulseand C_xis the capacitance of capacitor 41. Thus, for a switching frequency of f_s, the current I_CC1, fed into V_CCterminal by the discrete charges Q_x1is given by:

[0077] I_CC1=V_p·C_x·f_s  [Eq. 12]

[0078] C_xis charged to the voltage V_p−V_CC, according to the charge transfer. When V_pulsebecomes zero, the charge pump capacitor 41 (C_x) together with the inductor 42 (L_x) form a resonant circuit, which charges capacitor 41 (C_x) until the voltage across it reaches −V_CC. At that moment, the current of inductor 42 (L_x) flows into V_CCterminal until the current decays to zero. The energy transferred from capacitor 41 (C_x) to V_CCis given by: $\begin{array}{cc}\mathrm{Ecc2}=\frac{{\left(\mathrm{Vp}-\mathrm{Vcc}\right)}^2-{\mathrm{Vcc}}^2}{2}·\mathrm{Cx}& \left[\mathrm{Eq}.13\right]\end{array}$

[0079] and the current I_CC2delivered to V_CCis given by: $\begin{array}{cc}\mathrm{Icc2}=\frac{{\left(\mathrm{Vp}-\mathrm{Vcc}\right)}^2-{\mathrm{Vcc}}^2}{2·\mathrm{Vcc}}·\mathrm{Cx}·\mathrm{fs}& \left[\mathrm{Eq}.14\right]\end{array}$

[0080] From Eq. 13 and 14, the total current I_CCflowing into V_CCis given by: $\begin{array}{cc}\mathrm{Icc}=\mathrm{Icc1}+\mathrm{Icc2}=\left\{\mathrm{Vp}·\mathrm{Cx}+\frac{{\left(\mathrm{Vp}-\mathrm{Vcc}\right)}^2-{\mathrm{Vcc}}^2}{2·\mathrm{Vcc}}·\mathrm{Cx}\right\}·\mathrm{fs}& \left[\mathrm{Eq}.15\right]\end{array}$

[0081] The SCA resonant power supply described in FIG. 10, is of high efficiency, since most of the energy flowing in its reactive elements is recovered.

[0082] FIG. 11 illustrates an integration of the resonant power supply described in FIG. 10 into the SCA module, according to a preferred embodiment of the invention. Such integration improves the operation of the boost power stage, by providing a “soft switching” to switch 43 (S₁). Capacitor 44 (C_x) functions as a turn-off snubber for the switch, by slowing the rate of voltage developed across the switch. The overlapping time, when both voltage across the switch, and current through the switch are present, is reduced, reducing the switching losses. Moreover, since capacitor 44 (C_x) is charged (and discharged) by a current source (which is actually the inductor L_inin FIG. 1), the charging (and discharging) process is almost lossless, resulting in an improved efficiency.

[0083] According to another preferred embodiment of the present invention, a further improved APFC stage may be constructed, by adding a turn-on snubber to the circuit. This approach is illustrated in FIG. 12. The tapped inductor 45 (L_in), diodes 46 (Ds₁) and 47 (Ds₂), together with snubber elements (inductor 48 [L_r], and capacitor 49 [C_r]) reduce the switching losses during turn-on time. Still, the modular design presented in the invention is retained, even after adding the described snubbing circuitry or any other snubbing circuitry.

[0084] Production concept of SCA modules may vary according to the required quantities and power levels. FIG. 13 illustrates a construction of the APFC stage described in FIG. 6 above, according to a preferred embodiment of the invention. The SCA module is fabricated as a monolithic (chip) IC, packaged in a four pin package 50. This construction is suitable mainly for low power applications.

[0085] FIG. 14 illustrates construction of the APFC stage described in FIG. 7 above, according to another preferred embodiment of the invention. Here too, the SCA module is fabricated as a monolithic (chip) IC, packaged in a five pin package 51. This construction as well, is suitable mainly for low power applications.

[0086] FIG. 15 illustrates construction of the APFC stage, according to another preferred embodiment of the invention. The SCA module is fabricated as a multi-chip hybrid circuit, packaged in a three pin package 52, with an associate Printed Circuit Board (PCB) based sub-assembly 53. This construction is suitable mainly for medium and high power applications.

[0087] In case when the user prefers to use an (already) available power supply, an extra pin for external power supply may be easily provided, as shown in FIGS. 13 and 14 above. However, the main feature is related to minimal number of interconnections between the SCA module and other components.

[0088] FIG. 16 is a complete schematic diagram of an APFC stage structure, where the SCA module 78 includes internal power supply and turn-on snubber circuits, according to the invention, and is fabricated as a multi-chip hybrid circuit. The rectifier array is contained in a separate module 60. The input inductor 61 (L_in), which in this case is a tapped inductor, as well as the snubbing circuit inductor 62 (L_r), and the snubbing circuit capacitor 64 (C_r) are excluded from the SCA module, since they are not compatible with IC or hybrid technology. The control circuit controlling the switching transistor 77 (Q₁) is implemented using IC components. The voltage V_refis generated from V_CCby a resistor R₂and a zener diode D_z2, feeding the error amplifier 68 (U₃). The resulting error is fed into the multiplier 72 (U₄) together with the voltage across the sampling resistor 71 (R_s). The output voltage V_mfrom the multiplier is fed into comparator 66 (U₁), together with a ramp voltage V_ramp, which varies between a peak value of V_pand zero, thereby determining the switching frequency f_s(f_s=1/T_s).

[0089] The peak value of the ramp voltage V_pis determined from V_CCby resistor 75 (R_s) and zener diode 76 (D_z1). A current source 65 (U₅) charges the capacitor 69 (C_ramp) to V_ramp. V_rampis fed together with V_pinto another comparator 67 (U₂), which discharges Cramp by driving transistor 70 (Q₂) for a short determined period whenever V_rampexceeds V_m. In this way, the time T_offduring which the switching transistor 77 (Q₁) is not conducting, is proportional to V_m, and hence, to the input current i_ina(using Eq. 7 above). Using this construction, all the elements within the SCA module 78, ICs, semiconductors, resistors, capacitors and inductor 63 (L_x) may be integrated into one hybrid circuit.

[0090] A full APFC stage construction presented by FIG. 16 is illustrated in FIG. 17. Both rectifier array 80, and SCA 84 modules are ICs and/or hybrid circuits, packaged each in a heat dissipating package (Alumina, for instance) with metallic leads, and mounted on heat-sink 90 by drilling suitable holes 81 and 92 in the PCB 89, respectively, by gluing with heat conductive adhesive or soldering. The PCB 89 is mounted to the case by using screw(s) 91, or by soldering.

[0091] Printed copper traces 88 are used to connect between terminals of different components. The input tapped inductor 82 (L_in), the snubber inductor 83 (L_r), the snubber capacitor 86 (C_r), and the output capacitor 87 (C₀), are glued or soldered on the PCB material. Capacitor 93 (C₁) used for filtering high frequency components of the rectified voltage is optionally connected across the rectifier array 80 output. Each terminal of these components is soldered to a corresponding copper pad on the PCB.

[0092] Of course, the above examples and description has been provided only for the purpose of illustrations, and are not intended to limit the invention in any way. As will be appreciated by the skilled person, the invention can be carried out in a great variety of ways, employing more than one technique from those described above, all without exceeding the scope of the invention.