I. INTRODUCTION

Power quality has become an important consideration in power systems. Thus, much attention has been given to converters with high power factor (PF), low input harmonics, and good efficiency [1]-[3]. Power factor correction (PFC) provides many attractive benefits, including automatic line-voltage adjustments, nearly sinusoidal input current, and improved power quality. PFC front ends are now commonly used in power supply systems, which are connected to an AC line. This AC line ensures that power equipment are in accordance with many standards, such as IEC61000-3-2 specification [4], which is used to limit harmonic current; European Code of Conduct specifications [5], which defines the minimum average efficiency of external power supplies at 25%, 50%, 75%, and 100% of full load; and Energy Star [6], which sets the minimum limit at 0.9 PF.

In a step-down situation, one approach is two-stage converter: boost PFC front end [7], [8] combined with buck DC-DC downstream. PFC circuit is employed to obtain the sinusoidal input current to provide a high PF, whereas the DC-DC converter is used to regulate the output voltage. However, this structure has the disadvantages of low power density, high control complexity, and high cost because it includes additional PFC circuit. Single-stage [9]-[11] AC-DC converters have received much attention because of their cost effectiveness, compact size, and simple control mechanism. Another approach is buck-boost topologies, such as general buck-boost converter, fly-back converter [12], single-ended primary inductance converter (SEPIC), and Ćuk converter, which could act as step-down transformers [13]. In particular, SEPIC converter can achieve high PF and reduce output voltage stress. However, the voltage stress of switch is still relatively high. In addition, the efficiency decreases, and the cost increases.

Although buck converters are not widely used, these converters can also be used as a PFC front end [14]-[17]. Buck converter maintains high efficiency because of low average input current and root mean square (RMS) current at low input voltage. However, a buck PFC converter has difficulty achieving high PF because of its crossover distortion of input current when the input voltage is lower than output voltage. Moreover, the occurrence of input current distortion promotes total harmonic distortions (THD). Conventional buck PFC converter has difficulty passingIEC61000-3-2 standards because of THD.

An effective method to improve PF of the buck PFC converter is to modify the structure. Two step-down converters are proposed in [18], which include an auxiliary fly-back stage. The stage uses the same switch as a buck converter, and an additional power switch is commutated at the line frequency. The distortion of the input current can be eliminated with auxiliary fly-back converter. However, an additional diode that leads to additional losses is inserted in the power loop when these two topologies operate in buck mode. The interleaved power converter that operates at the boundary conduction mode (BCM), which is proposed in [19], can be used to recover the distortion of input current. However, the additional coupled inductor complicates the structure, which makes this structure unsuitable for such an arrangement.

In this study, a pulse-assisting buck (PAB) PFC converter is designed. The envelope of input current of the proposed buck converter is quasi-sinusoidal when buck-boost topology is turned in discontinuous mode where the drive signal is provided by additional pulse during distortion time. The predominant drawback of buck PFC can be overcome by adding discontinuous conduction mode (DCM) operation during the distortion time to make the current continuous. The DCM operation of all single-switch topologies is most suitable for low-power applications [20]. The proposed PAB converter is suitable for step-down PFC front end of an AC-DC converter.

The detailed operation principle is illustrated in Section II, the circuit parameters design considerations are presented in Section III, and the results based on a 94 W prototype are shown in Section IV. A conclusion is presented in Section V.

II. OPERATION PRINCIPLE

The output bus voltage of a conventional buck converter that is connected to a 50HzACsource and bridge rectifier is set at a level less than the peak AC voltage at the lowest line voltage. When the instantaneous AC input voltage is greater than the output bus voltage, the PFC stage is forward biased, and the current can be drawn from the AC input. However, when the instantaneous AC input voltage falls below the bus voltage level, the diode bridge rectifiers become reverse biased, and no power can be drawn from the AC line. This condition is called crossover distortion of the input current, as shown in Fig. 1. The conduction angle can be calculated from (1).

Fig. 1.Crossover distortion of the input current.

A. Preliminary

For the proposed PAB converter, the PAB converter operates in BCM, which is same as the conventional buck converter, when the instantaneous AC input voltage is higher than the output voltage [22]. When the input voltage is lower than the output voltage, the PAB converter operates in DCM, in which the gate drive voltage is provided by the PFC controller. Hence, the PF can be improved obviously and THD can be substantially eliminated based on preserving the merits of the conventional buck converter.

The current waveforms of the PAB converter are shown in Fig. 2. When the PAB buck converter operates in BCM, the positive edge of gate drive voltage is determined by zero current detection (ZCD) of the inductor current. When the converter operates in DCM during the distortion time, the turn-on instant is determined by the internal oscillator of PFC controller. The conduction of switch Q1 and Q2 in Fig. 3 is initiated by the oscillator of the controller in every beginning of a period.

Fig. 2.Current waveform of the proposed converter.

Fig. 3.Schematic of the PAB converter and its controller.

B. Working Process

The converter operates in several stages during a sinusoidal period, as shown in Fig. 4.

Fig. 4.Operating stages of the PAB converter.

1) DCM: When the magnitude of input voltage Vin is smaller than the output voltage Vo, the proposed converter operates in DCM. The switches Q1 and Q2 are switched on and off synchronously and controlled by internal pulse.

When Q1 and Q2 turn on, the proposed converter operates in Stage 1. The equivalent circuit is illustrated in Fig. 4(a). In this stage, inductor L is charged by Vin, and inductive current iL increases during this stage.

When Q1 and Q2 turn off, the proposed converter operates in Stage 2. The equivalent circuit is illustrated in Fig. 4(b). In this stage, inductor L is discharged by Vo, and iL decreases during this stage.

2) BCM: When the magnitude of Vin is greater than Vo, the proposed converter operates in BCM. The switch Q2 remains switched off, and switch Q1 is controlled by signal PRI by the PFC controller.

When Q1turns on, the proposed converter operates in Stage 3. The equivalent circuit is illustrated in Fig. 4(c). In this stage, inductor L is charged by Vin−Vo, and iL increases during this stage.

When Q1turns off, the proposed converter operates in Stage 4. The equivalent circuit is illustrated in Fig. 4(d). In this stage, inductor L is discharged by Vo, and iL decreases during this stage.

Based on the preceding analysis, the PFC controller is designed as shown in Fig. 3.

The input voltage Vin is detected and divided to obtain a signal MULT. Similarly, the output voltage Vo is represented by FB with the same ratio as that of MULT representing Vin. Then, the control signal Vp can be calculated by comparing MULT and FB. The signal Vp is high logic when MULT is higher than FB and is low logic when MULT is lower than FB. The control signal Vpis used to determine whether the converter operates in DCM or BCM.

The detected output signal INV is sent to the negative input of the error amplifier. The error between INV and the set reference voltage Vref is amplified by the compensation networks, and then the amplified error signal Ver is achieved. The DC voltage signal Ver is applied to one input port of the multiplier, whereas another input signal of the multiplier is MULT. The waveform envelope of the output of the multiplier is a rectified sine wave. The output of the multiplier is used as a reference signal of the current comparator, and the output of the current comparator is used to control the peak current per cycle of the switch. When the voltage of the buck inductor is reversed, ZCD module opens the external switch. Thus, the circuit operates in BCM.

Two of the input signals of the selectors are the internal pulse and the output of the RS trigger, whereas other input signals of the selectors are the internal pulse and ground. The output signals PRI and SEC of the two selectors are the gate signal of Q1 and Q2, respectively.

The PAB converter operates in DCM controlled by an internal oscillator. The frequency of the internal pulse is fs, and the duty cycle is D0 during distortion time. The PAB converter operates in boundary mode during conduction time by using constant on-time (COT) control. The waveforms of nodes MULT, FB, Vp, PRI, and SEC and of input current iin are shown in Fig. 5.

Fig. 5.Key waveforms of nodes MULT, FB, Vp, PRI, and SEC and iinof PAB converter.

The signal Vp determines different operation modes alternately per cycle by controlling the gate signals PRI and SEC. Different frequencies fs of internal pulse result in the different performances of the PAB converter. The detailed optimization of frequency fs and duty cycle D0 during the distortion time is presented in Section III.

III. ANALYSIS OF THE PAB CONVERTER

Design considerations of the PAB converter controlled by the proposed PFC controller are presented in this section. All the calculations are based on a 94Wprototypewith 90V to 264V universal AC input. The output voltage is set to 80V. The circuit described in this paper can easily obtain high PF. The following are the analyses of the input current and how the frequency and the duty cycle of internal pulse during the distortion time affect the performance of the circuit.

A. Current Analysis

According to the operating stage analysis shown in Fig. 3, the characteristics of the PAB PFC converter can be derived. When the input voltage is lower than the output voltage in a half line cycle, the converter is equal to a DCM buck-boost converter. When the input voltage is higher than the output voltage, the PAB converter works like a BCM conventional buck converter. The input current in the half line cycle can be described as follows, whereas the conduct angle can be calculated from (1)

where DBCM and DDCM are duty cycles in BCM and DCM, respectively. The peak value of the input current is ipk,BCM and ipk,DCM in BCM and DCM, respectively.

The following is the final expression of the input current obtained by substituting (3) and (4) into (2):

where ton is the on-time of the primary switchQ2, which is nearly constant during the half line cycle. We introduce a factor, and replace D02/ fs by k*ton to simplify the calculation.

We rewrite the expression of the input current as follows:

According to the law of energy conservation, we can obtain the equation as follows:

where ton can be achieved as follows:

Moreover,

According to (6), the calculated input current in one cycle with different factor k at 220V AC input level is shown in Fig. 6. The results and discussion are shown in Section IV.

Fig. 6.Calculated input current with different factor k at 220V AC input.

The preceding analysis indicates that the input current varies from different k and only the coefficient k in a certain range could enable the converter to meet the requirements. These conditions occur because the coefficient k is the function of D0 and fs. The selection of D0 and fs is analyzed in the next section.

B. PF Analysis and THD Analysis

A converter that maintains PF for over 0.9 must be designed in the same way as Energy Star does, which introduces a PF legislation of 0.9 at full load. From the preceding calculation, we can obtain the expression of PF as follows:

where irms is the RMS value of the input current that can be achieved as follows:

Moreover, Pi is the input power in a cycle

The following expression of PF and THD are obtained by substituting (11) and (12) into (10):

According to (13)and(14), the corresponding values of PF and THD with different k at 220V and 50Hz AC input level are shown in Table I. The trend of PF and THD versus k is shown in Fig. 7.

TABLE ICALCULATED PF AND THD VERSUS k AT Vin = 220V AC INPUT COMPARED WITH IEC61000

Fig. 7.(a) Trend of PF versus k. (b) Trend of THD versus k.

When k=0, the PAB converter works as a conventional buck converter. Fig. 7 and Table I demonstrate that with the increase of k, PF increases and THD decreases. However, the value of k could not increase without limit in consideration of the distortion of the input current in a relatively large k.

The tons for different values of k are also included in Table II. ton decreases gradually with the increase of k. In fact, ton is nearly equal to the ton of the conventional buck converter only in a case of small values of k, that is, below 0.1.

TABLE IIRELATIONSHIP BETWEEN k AND ton

A small k (i.e., below 0.1) is a good choice to preserve the superiority of the buck converter and to consider some factors, such as PF and THD.

In this paper, ton has a one-to-one correspondence with k. Thus, these two variables can be used as one constraint condition. According to the function D02/ fs = k*ton, D0 and fs are two independent parameters. Moreover, if one of them must be ensured, another must be known. For small k, we assume that the value of ton is equal to that of the conventional buck converter according to the preceding analysis. For the buck converter without PAB technology, we can obtain the switching frequency function as follows:

where

Thus,

If the output voltage is designed at 80 V, the buck inductance is obtained as L = 150 μH. Fig. 8 shows the calculated switching frequency versus the function of the phase angle in 220V and 50Hz AC input voltage process.

Fig. 8.Calculated switchingfrequency versus t of conventional buck converter.

In this paper, k can be used to characterize the power transmission capacity of the circuit. Each value of k may correspond to a specific combination of different D0 and fs, as shown in Fig. 9.

Fig. 9.Relationship between fs and k, D0 at 220V AC input.

The range of operating frequency illustrated in Fig. 8 is approximately 0.1MHz to 0.27Mhz. Therefore, Fig. 9 does not show the situation where k is greater than 0.1 because when k is greater than 0.1, fs gradually decreases to a value smaller than the minimum value shown in Fig. 8.

The selection of D0 and fs must not only meet requirements of the operating frequency of the circuit but also consider the output level of voltage and power. In view of these two factors, the coefficient k is set to 0.08 according to Fig. 9, and fs is set to the lowest frequency (0.1Mhz)in Fig. 8. D0 is set to 0.2as a result of the function D02/ fs = k*ton. Then, the validation is based on these values, which proves that this solution is one of the feasible solutions.

IV. RESULTS AND DISCUSSION

The model derived from the previous section uses a 94W, 220 V AC input, 80V DC output PAB PFC converter as an example. Moreover, the converter can be used as the universal input (Vin = 90Vto 264 V) front end.

Input voltage and input current waveforms of the conventional buck PFC converter are shown in Fig. 10. Input voltage and input current waveforms of the PAB PFC converter are shown in Fig. 11. We can see that the strategy restores the input current during distortion time.

Fig. 10.Input voltage and input current waveforms of the conventional buck PFC converter.

Fig. 11.Input voltage and input current waveforms of the PAB PFC converter.

The PF and efficiency from 90V input to 264V input of the PAB converter are listed in Table III. The buck PFC maintains a high efficiency of approximately96% when the input voltage is larger than 115V. The slight drop of efficiency is the result of the implementation of the control circuit with discrete ICs. Thus far, no dedicated controller IC for the buck PFC is available.

TABLE IIIPF AND EFFICIENCY FROM 90V INPUT TO 264V AC INPUT OF THE PAB CONVERTER

Efficiency versus load at 110 V AC input and 230 V AC input is presented in Table IV. Table IV demonstrates that the buck PFC maintains a high efficiency across the load range from 100% to 25% of the full load.

TABLE IVEFFICIENCY VERSUS LOAD AT 110V AC INPUT AND 230V AC INPUT

The efficiency and PF of the PAB PFC converter is compared with those of a previously proposed converter of the same type as [21] in Fig. 12. The PAB converter has a higher PF particularly under low input line voltage and high line voltage conditions. Moreover, the efficiency of the PAB converter has a good performance under high input line voltage. The efficiency of the PAB converter is slightly deteriorated compared with that of the prior converter under low input line voltage conditions. The added two diodes and a switch contribute to the total loss.

Fig. 12.(a) Measured PF of PAB converter. (b) Measured efficiency of PAB converter.

Fig. 13 shows the THD performance comparison between the PAB PFC converter and the conventional buck converter, which has the same parameters with those of the PAB converter but without pulse-assisting technique.

Fig. 13.THD performance comparison between the PAB PFC converter and the conventional buck converter. (a) Iac(n) compared with IEC(n) of PAB converter at 110VACinput. (b) Iac(n) compared with IEC(n) of conventional converter at 110VACinput. (c) Iac(n) compared with IEC(n) of PAB converter at 230V AC input. (d)Iac(n) compared with IEC(n) of conventional converter at 230V AC input.

To compare with IEC61000-3-2 Class C limits, the vertical axis of Fig. 13 is set as the value of Iac(n), where Iac(n) is the n-order harmonic content of the input current iin, and IEC(n) represents the n-order harmonic current limit of IEC61000-3-2 Class C. For example, 110V AC input voltage and 230V AC input voltage are needed to represent normal low line and normal high line, respectively. The harmonic performance of the PAB converter meets the IEC61000-3-2 Class C requirement under both voltages, whereas the conventional buck without pulse-assisting technique cannot meet the requirement of both voltages. In addition, the PAB converter performs effectively under normal high line.

V. CONCLUSION

According to the design considerations given in the previous sections, a 94W, 80V output, universal input (Vin=90 V to 264 V) quasi COT-controlled buck PFC prototype is produced. The IC is based on L6561 with the addition of an oscillator, a voltage subdivision structure, a comparator, and two selectors to control the PAB converter. The improved PF and THD can approach the ideal values compared with the calculated PF and THD under different selections of k factor.

The PAB PFC maintains PF of approximately98% at nominal high line and approximately96% at nominal low line. In addition, the PAB PFC maintains an efficiency of 98% at nominal high line and approximately96% at nominal low line. The PAB converter achieves high efficiency across a load range from 100% to 25% of full load, and the input current harmonics can meet the IEC61000-3-2 Class C standard within the universal input voltage range. In addition, according to the preceding data, the PAB converter performs well at normal high line voltage, but it could meet the requirements at normal low line voltage.