1. Introduction
With the rapid advancement of electronic devices, the demand for switching power supplies has surged significantly. Power factor correction (PFC) converters have been extensively utilized to mitigate input current distortion and ensure compliance with stringent harmonic standards. Certain commercial power supply products must comply with the IEC 61000-3-2 standard to mitigate the potential damage to the power grid caused by harmonic injection [
1,
2,
3,
4,
5].
Among the various PFC converter topologies commonly used, the Boost converter is particularly favored because of its ability to achieve low input current ripple and maintain high efficiency. However, the Boost PFC converter faces several challenges, including high output voltage due to the need for the output voltage to exceed the peak input voltage, large input inrush current, and difficulty in achieving output short-circuit protection. These limitations restrict the applicability of the Boost PFC converter in certain scenarios [
2,
3,
4]. The conventional Buck PFC converter offers advantages such as low voltage stress on the power switch and effective step-down conversion. Nevertheless, its operation is limited to lower voltage conditions. When the input voltage falls below the output voltage, a dead zone in the input current occurs, leading to a degraded power factor (PF) and increased total harmonic distortion (THD) of the input current. This restricted output voltage range further limits the applicability of the Buck PFC converter [
5,
6,
7]. The Buck-Boost PFC converter offers inherent current shaping, cost-effectiveness, and the capability to perform both step-up and step-down conversions. However, compared to the traditional Boost and Buck converters, Buck-Boost PFC presents several drawbacks including low efficiency and high voltage stress of the main power switch due to the inductor only supplying energy to the load when the main switch is off, and the discontinuous input current leading to worse harmonic distortion in the power grid [
8,
9].
The Cuk converter, introduced by Slobodan Cuk in 1977 [
10], features an input stage similar to a Boost converter, which provides the advantage of continuous input current and can be used for achieving PFC. The output stage of the Cuk converter resembles a Buck converter, enabling DC-DC conversion. Consequently, the Cuk converter generates minimal electromagnetic interference and output voltage ripple, which seems to be a better candidate in the basic PFC converter topology [
11], leading to its widespread application in recent years. In [
12], a Cuk PFC converter is used for driving brushless DC motors, which reduces system power consumption, achieves high efficiency, and also lowers costs. In [
13], to improve the efficiency of the two-stage converters for electric bike battery charging, a switched inductor Cuk PFC converter was proposed. Compared to conventional two-stage converters, this topology offers advantages such as higher voltage gain, reduced current stress, and improved efficiency. In [
14], a three-phase-isolated Cuk converter-based PFC rectifier is proposed. The converter operates in discontinuous output inductor current mode and requires only a simple voltage control loop to achieve PFC for the AC input. This design reduces the system cost while enhancing reliability and efficiency. Therefore, the Cuk PFC converter is widely used in many applications, but it still has some shortcomings.
In [
15], the harmonic balance method and Floquet theory are used to study the influence of intermediate capacitance on the slow-scale instability of operating in discontinuous capacitor voltage mode with a Cuk PFC converter. The research results show that with the increase in intermediate capacitance, the stability of the converter will be reduced due to period-doubling bifurcation, thereby reducing PF. The bridgeless Cuk PFC converter, by reducing the number of semiconductors in the circuit, lowers conduction losses and finds widespread application in various fields. However, it is challenging for the converter to achieve both high PF and low intermediate capacitor voltage simultaneously during normal operation and the converter can only operate under low input voltage [
16,
17,
18]. In [
19], a Cuk PFC converter utilizing a variable inductor is proposed, which makes both output and input inductors operate in DCM. The implementation of variable inductor (VI) control effectively enhances the PF of the converter. However, at high input voltages, the peak current in the output inductor becomes significantly large, which increases the current stress on the switch, ultimately reducing the efficiency of the converter.
Variable inductor technology has changed the characteristics of inductor components, transforming the inductor from a fixed inductance component to a variable inductance component. VI technology can add an extra control variable to the switching power supply circuit, allowing the switching power supply to change from only adjusting the duty cycle to simultaneously adjusting both the duty cycle and power inductance. Therefore, VI technology has attracted considerable interest from many researchers.
Recent advancements in science and technology have spurred rapid development in magnetic materials, leading researchers to propose various types of variable inductors. Commonly used variable inductors include DC auxiliary winding types such as the toroid core VI, double-E core VI, and quad-U core VI [
20,
21,
22,
23,
24]. These designs utilize DC magnetic flux to drive core saturation, thereby altering the core equivalent magnetic permeability to adjust the inductance value.
A variable inductor offers inherent flexibility of the inductance variation, which allows converters to handle more complex operating conditions. Variable inductors have been extensively applied in many cases such as PFC converters, electric vehicles, electronic ballasts, and LED drivers [
24,
25,
26,
27,
28,
29,
30]. In order to improve the current handling ability and reduce the size of the magnetic component of the power inductor, a toroid core variable inductor, which enhances the inductance value by de-saturating the core with DC magnetic flux generated by a direct current, is proposed to replace the conventional power inductor in the bidirectional DC-DC converter in [
24]. In [
25], a double-E core type variable inductor is employed in a critical conduction mode (CRM) Boost PFC circuit proposed to change the inductance value, thereby increasing the switching period, minimizing switch turn-off losses, and enhancing system efficiency.
In [
26], a new analog control method is utilized to improve the PF of Boost PFC operated at CRM. It has been pointed out that the poor PF is due to the input current phase lead caused by the presence of an input capacitor. It can increase PF by producing a lagging current utilizing a variable inductor to compensate for the input current phase lead. In [
28], the power of the fluorescent lamp is adjusted by controlling the inductance value to achieve the purpose of dimming, thereby enhancing the linear dimming range and achieving high efficiency and high PF. A unidirectional resonant switched-capacitor step-up converter for OLED driving is proposed in [
29]. This converter combines a variable inductor with a capacitor in series to control the OLED current and adjust the brightness.
Variable inductor technology is a technique that can change the magnitude of inductance by controlling the saturation level of the inductance core to operate the inductance in a nonlinear region, thereby changing the magnetic permeability of the magnetic circuit and achieving changes in inductance value. In recent years, some scholars have also conducted research on inductors working in nonlinear regions in switch-mode power supplies and proposed corresponding analysis methods to better understand their working characteristics and their effect on converter systems.
A method for estimating the current flowing through nonlinear power inductors in DC/DC converters was proposed in [
31] and applied to Boost converters. The proposed method uses a polynomial third-order inductor model to estimate the current distribution at which the inductor reaches saturation, providing a complete evaluation of the current, including distribution, peak value, root mean square value, and spectrum. In [
32], a dedicated measurement rig was developed to reflect the behavior of inductance during saturation and temperature rise. And it is pointed out that the saturation current and loss of the inductor decrease and increase respectively with the increase in temperature.
To achieve a high power factor while maintaining a low intermediate capacitor voltage and ensuring high efficiency with wide input voltages, by operating the input inductor at DCM and the output inductor at CRM, a DCM-CRM Cuk PFC converter utilizing variable inductor control is proposed in this paper, wherein the input inductance value is precisely regulated by DC magnetic flux bias. According to [
33,
34,
35], when the inductor in a Boost converter operates in DCM, the converter can inherently realize PFC without the need for additional control circuitry. Furthermore, studies in [
36,
37] indicate that a Buck converter operating in CRM exhibits higher efficiency compared to its operation in DCM. Given that the input stage of a Cuk converter is analogous to a Boost converter and its output stage mirrors a Buck converter, the Cuk converter inherently benefits from these advantageous characteristics. Therefore, this paper configures the input inductor to operate in DCM and the output inductor in CRM.
A detailed analysis of the converter topological principle, control strategy, and operational characteristics is provided in this paper. To validate the probability of the proposed approach, a 108 W experimental prototype was built and tested. The test results confirmed the accuracy of the theoretical analysis and demonstrated the effectiveness of the proposed control method. The experimental results demonstrate that the proposed converter not only reduces the intermediate capacitor voltage but also enhances the PF and improves efficiency with wide input voltages.
This paper is organized into 5 sections. In
Section 2, the characteristics and operating principles of the conventional DCM-CRM Cuk PFC converter are presented. In
Section 3, the proposed DCM-CRM Cuk PFC converter with VI control is introduced, including a detailed explanation of its operational principles and analysis of key features. A comprehensive comparison of experimental results of the conventional and proposed Cuk PFC converters is also provided in
Section 4, and the conclusions are summarized in
Section 5.
2. Operation Principle of the Conventional DCM-CRM Cuk PFC
The main circuit configuration of the conventional DCM-CRM Cuk PFC converter is depicted in
Figure 1. As illustrated, the main components include a rectifier bridge, an input LC filter
Lf and C
f, two power diodes
D0 and
D1, a power switch
S1, an input inductor
L1, an output inductor
L2 with auxiliary winding for zero current detection, an intermediate capacitor
C1, and an output capacitor
C2.
The operating waveforms of the conventional DCM-CRM Cuk PFC converter are depicted in
Figure 2. From top to bottom, the waveforms illustrate the switching drive signal
Vg, the input inductor current
iL1, and the output inductor current
iL2. The DCM-CRM Cuk PFC converter undergoes three distinct operating modes within each switching cycle.
Mode I (0–
t0): During this interval, the drive signal
Vg activates the switch, causing the voltage at the left terminal of the intermediate storage capacitor
C1 to drop rapidly to zero. And since the voltage at the two ends of the capacitor cannot be changed abruptly, the voltage at the right end of the capacitor becomes negative. This results in the diode
D1 being subjected to reverse voltage, thereby turning it off. The AC power source charges the input inductor
L1 through diode
D0 and switch
S1. Simultaneously, the intermediate energy storage capacitor
C1 powers the output inductor
L2, output capacitor
C2, and load via the
S1. The increasing slopes of the currents flowing through the input and output inductors can be expressed as
where
vrec denotes the rectified voltage;
VC1 represents the voltage across the intermediate capacitor
C1; and
Vo indicates the output voltage.
Mode II (
t0–
t1): In this mode, the driving signal
Vg turns off the switch, while diode
D0 remains in conduction. Diode
D1 is subjected to a positive voltage, initiating conduction to provide access to the input circuit. The rectified voltage
vrec and the input inductor
L1 commence charging the intermediate capacitor
C1 through
D0 and
D1. Meanwhile, the output capacitor
C2 stores energy transferred from the output inductor
L2. The rate of voltage decrease across both input and output inductors can be expressed as
Mode III (t1–t2): The input inductor current iL1 drops to zero, causing diode D0 to turn off, while diode D1 remains conducting. During this phase, the output inductor L2 supplies the load through D1, forming the output circuit, and the current iL2 continues to decrease. When the output inductor current iL2 reaches zero, it is detected by the zero-crossing detection (ZCD) module. This triggers the driving signal Vg to turn on the switch, transitioning L2 to CRM, and initiating the next operating cycle of the converter.
According to Equation (1), during one switching cycle, the peak current
iL1_pk(
t) through the input inductor
L1 can be expressed as
where
ton is the turn-on time of switch
S1 within one switching period, it is critical for determining the peak current
iL1_pk(
t) through the input inductor
L1.
VM is the amplitude value of the AC input voltage.
ω is the angular frequency of the AC input voltage. The discharge time
toff1 of
iL1 can be derived from the volt–second balance of the input inductor
L1.
Thus, based on (5) and (6), the input current
iin(
t) is represented by the average current
iL1(
t) flowing through the input inductor during the switching cycle. Consequently, the input current can be given as
where
Ts is the switching period and
toff1 is the time for the input inductor current to drop to zero. According to (2), the peak current
iL2_pk(
t) of the output inductor
L2 can be expressed as
The average output inductor current is equivalent to the output current; hence, the output current
Io can be expressed as
where
toff2 is the discharge time for the output inductor current to drop to zero. Based on the volt–second balance of inductor
L2, the
L2 discharge time
toff2 formula can be expressed as
As shown in
Figure 2, the output inductor operates in CRM mode, and from Equation (9), the peak current
iL2_pk(
t) of the inductor
L2 can also be expressed as
Substituting Equation (10) into Equation (9) yields an expression for the conduction time
ton of the switch as
Neglecting the voltage ripple of the intermediate capacitor, Equations (8) and (11) indicate that, when the converter is delivering energy to the load, the conduction time
ton of
S1 can also be expressed as
When the switch is turned off, by associating Equations (10) and (13), the turn-off time
Toff of switch
S1 is obtained as
From (5), it can be seen that the discharge time of the inductor
L1 is the longest when
iL1_pk(
t) reaches the maximum at |sin(
ωt)| = 1 in a half-line cycle, and the longest discharge time
toffmax can be expressed as
Inductor
L1 can be operated in DCM mode for the entire frequency cycle by making it operate in DCM mode at the peak input voltage. The conditions for the inductor
L1 to operate in DCM mode can be expressed as
By associating (15) and (16), the condition under which the inductor
L1 works in DCM can be obtained as
Without considering the converter losses, it can be obtained based on the conservation of input power and output power:
According to (7) and (12), Equation (18) can be rewritten as
where
K =
L2/L1. Based on Equation (7), the expression of PF can be expressed as
According to (19), the magnitude of
VC1 can be computed using an iterative method once the load parameters are established.
Figure 3 illustrates the variations in the PF and the intermediate capacitor voltage as functions of the effective input voltage
Vin_RMS for different ratios of
K. From
Figure 3, it is evident that, for a constant
K,
VC1 increases with the increase in
Vin_RMS. Concurrently, within the range of 90 Vac to 240 Vac, the PF experiences a slight decrease as
Vin_RMS rises. Additionally, for a given
Vin_RMS,
VC1 rises with an increase in the ratio
K, while the PF diminishes as
K decreases.
The critical value of the
K value when the inductor operates in DCM mode under the entire AC input voltage range and given load can be determined by the limiting condition Equation (17) for the input inductor operating in DCM mode and the relationship Equation (19) for the intermediate capacitor voltage under the conventional control method. When
K equals or exceeds this threshold, the inductor operates in DCM mode; if
K is below this threshold, the inductor operates in CCM mode. Therefore, the variation curve can be obtained from the obtained
K value and Equation (19), as shown in
Figure 3. It can be seen that
VC1 is only related to the value of
K under a certain load condition. In the region of
K < 1.4,
L1 operates in CCM mode, and in the region of
K ≥ 1.4,
L1 operates in DCM mode.
To ensure that the input inductor
L1 operates within the DCM mode constraints, and based on the analysis from
Figure 3, it was observed that when 1.4 <
K< 3.6, the PF remains above 0.96. As the input voltage is fixed, the intermediate capacitor voltage
VC1 increases with the rise in the
K value, necessitating a higher withstanding voltage rating for the intermediate storage capacitor
C1.
According to the above derivation of the VC1 and PF value of the conventional converter, it can be known that both the PF and the intermediate capacitor voltage are influenced by the ratio K = L2/L1. An increase in the ratio K leads to a higher PF, however, it simultaneously results in an elevation of the VC1. Based on the comprehensive comparison of PF and the intermediate capacitor voltage, K = 2.05 was selected in this paper. However, Capacitors rated above 450 V are expensive and offer lower cost-effectiveness. Consequently, achieving both a high PF and a low intermediate capacitor voltage becomes challenging. To address this problem, a DCM-CRM Cuk PFC converter based on variable inductor control which dynamically adjusts the input inductor in response to variations in the transient rectified input voltage is proposed in this paper. Since both the PF and VC1 are related to the ratio K, it is possible to change the inductance of L1 during the half-line cycle and change the ratio K in real time. In contrast to the conventional Cuk PFC converter, the converter proposed in this paper utilizes a variable inductor as the input inductor. The inductance is dynamically adjusted according to a specific control strategy to achieve a high power factor while effectively reducing the voltage of the intermediate capacitor. Concurrently, the output inductor current remains constant, leading to a reduction in peak inductor current compared to the approach. This decrease in current stress and switching losses subsequently enhances the overall efficiency of the converter.
4. Experimental Results
To verify the correctness of the analytical results of the proposed DCM-CRM Cuk PFC converter utilizing variable inductor control, a 108 W experimental prototype was constructed. Comparative experimental results between the conventional and proposed converters were obtained. Detailed specifications and key circuit parameters are provided in
Table 1.
Figure 14 and
Figure 15 depict the experimental prototype and the experimental setup respectively.
Figure 14a–c show the main board, control board, and digital control board for variable inductor calculation of the experimental prototype. As the control board shares the same control board with other CRM isolated converter prototypes, some components on the control board were not soldered when used for this DCM-CRM Cuk PFC converter. The digital control board for variable inductor calculation in
Figure 14c uses STM32F103RCT6 as the controller, the main program code for variable inductor calculation unit is shown in
Appendix A.
The experimental results for the input voltage
Vin, input current
iin, intermediate capacitor voltage
Vc1, and output voltage
Vo of both the conventional Cuk PFC converter and the proposed converter are presented in
Figure 16. These figures illustrate the performance of the converters at input voltages of 110 Vac and 220 Vac, respectively.
As illustrated in
Figure 16a–d, the output voltage of both the conventional Cuk PFC converter and the proposed converter is stable at 72 V. However, under conventional control, the input current exhibits harmonic distortion at 110 V and 220 V input voltage. In contrast, the proposed converter effectively mitigates this harmonic distortion, the input current of the proposed converter is closer to sinusoid wave. Furthermore, the intermediate capacitor voltage of the conventional converter is around 270 V and 500 V at 110 V and 220 V input voltage, respectively; the intermediate capacitor voltage of the proposed converter is reduced to around 230 V and 400 V at 110 V and 220 V input voltage, respectively. This demonstrates that the proposed converter effectively reduces the intermediate capacitor voltage.
The experimental waveforms for the input inductor current and output inductor current at 110 V and 220 V input voltage are presented in
Figure 17 and
Figure 18. From
Figure 17, it is noted that the peak current of the input inductor is reduced from 6 A in the conventional converter to 4.8 A in the proposed converter at 110 V input voltage. From
Figure 18, it is illustrated that, at 220 Vac, the peak current value of the input inductor decreases from 5 A of the conventional converter to 3.4 A of the proposed converter. The peak output inductor current remains consistently at 3 A. Consequently, the proposed converter effectively reduces the current stress on the switch and enhances the overall efficiency of the converter, which is the same as the theoretical analysis result.
As depicted in
Figure 17d and
Figure 18d, the current through the variable inductor
iLV exhibits slight distortion. This distortion arises because the inductor design includes a margin for variation, and the actual operational range of the variable inductor is narrower than anticipated. Consequently, a stable DC current must be continuously applied to the bias winding to maintain the calculated inductance value. When the DC current is applied, the operating point of the magnetic core characteristic curve shifts toward the saturation region [
30], leading to a reduction in the inductance during actual operation and resulting in a curved waveform for the inductor current. This shift results in the actual inductance value of the variable inductor being lower than the theoretical calculation, thereby causing the inductor current to exceed the theoretically predicted value.
Figure 19 shows the experimental data of the PF and efficiency. It is obvious that the PF of the proposed converters is improved in the range of input voltage variation, from 0.986 to 0.998 at 110 Vac and from 0.981 to 0.992 at 220 Vac. As illustrated in
Figure 19, the efficiency is markedly enhanced with the use of the variable inductor, achieving values exceeding 90% at 90 V input voltage. At the same time, the efficiency of the proposed converter reached 86.8% which is higher than 81.1% of the conventional converter at 240 V input voltage. It can be seen that compared to the conventional converter, the PF and efficiency of the proposed converter are obviously enhanced.
The test data of input current harmonic content and THD of the conventional and proposed converter are compared in
Figure 20. It can be seen that compared with the conventional one, the proposed converter can more easily meet the requirements of IEC61000-3-2 class D. The proposed converter greatly reduces THD and the harmonic current content, especially the 3rd harmonic current.
From
Figure 5, it can be seen that the main power losses of the conventional and the proposed converter include the losses of the rectifier bridge, input filter, input inductor
LV or
L1, a power switch
S1, a diode
D0, a free-wheeling diode
D1, and an output inductor
L2. The power loss distribution diagram is presented in
Figure 21. These figures illustrate the power loss of the converters at input voltages of 110 Vac and 220 Vac, respectively. From the figure, it can be seen that the proposed converter mainly reduces the losses of the power switch and input inductor, thereby increasing the efficiency of the converter.
Table 2 shows that compared to the conventional DCM-DCM Cuk PFC converter, which is mentioned in [
19], and the traditional DCM-CRM Cuk PFC converter, the proposed DCM-CRM Cuk PFC converter enhances the PF, reduces the THD, decreases the intermediate capacitor voltage, and improves efficiency.