High Step-Up Converter

Abstract

A new high-boost converter derived from a charge-pump capacitor-based dual-boost converter is presented herein. Compared with the existing charge-pump capacitor-based dual-boost inductor converter, this structure adds an additional voltage-doubling circuit and additional coil to the original circuit to increase the voltage conversion ratio and winds the three coils around the same magnetic element. By this way, the overall size of the circuit can be reduced, and the leakage energy of the coupled inductor can be recovered. In this paper, the proposed new high-boost converter is firstly explained and analyzed, followed by simulations to demonstrate the feasibility of the proposed circuit and finally the experimental results to verify its effectiveness.

1. Introduction

Nowadays, most countries need to adopt large-scale construction of power generation methods, such as thermal power generation, nuclear power generation, and hydroelectric power generation, in order to obtain the electricity needed for daily life. Although these methods have the advantage of obtaining lower electricity prices, they are often limited by location. For example, thermal power generation needs to be located far away from cities to prevent air pollution, hydroelectric power generation needs to be located in places where there is a source of water, and nuclear power generation has to be located at the seacoast to use seawater for cooling. In the past, the cost of setting up these large-scale power plants was very high, but in recent years, due to the improvement of power generation reliability, the cost of setting up the plants has been lower than the cost of maintaining the power grid, so many countries have invested in the research and development of microgrids and decentralized power generation technology. There are several types of decentralized power generation, such as waste heat, solar power, grid-supplied electricity from electric vehicles, and wind power. However, the voltages generated by the above power generation methods are not high and stable enough, and in addition, the output voltage is prone to change with the change of load. To solve this problem, a step-up converter is connected to the back end of the decentralized energy source to provide a relatively stable and high voltage for the back-end equipment or to provide the input power for the inverter of the utility paralleling system, which is needed for the power supply. For example, in the thermoelectric generator (TEG), stand-alone solar power system [1], and utility-parallel solar power system [2], a step-up converter is required to step up the low voltage generated by the thermoelectric modules or solar panels to a high voltage to provide power to the inverter or load.

Generally speaking, the voltage conversion ratio of a conventional boost converter is theoretically infinite, but due to the parasitic components, it can only be increased to about four times [3]. Therefore, many novel boost converters have been proposed, such as the KY converter and its related derivatives [4,5,6,7], which operate on the principle of obtaining a high voltage conversion ratio by means of a simple charge-pump circuit with a boost inductor or coupled inductor. In the literature [8,9,10,11,12], charge-pump circuits with single or multiple magnetic components are used to improve the voltage conversion ratio. The literature [13,14,15,16] proposes interleaved boost circuits to reduce the relatively large input currents; the main principle is also the use of coupled inductors to achieve high voltage conversion ratios. The literature [13] presents a variant of the interleaved boost converter, which is characterized by bidirectional energy conversion. The disadvantage is that the number of controllable switching elements is too large and hence the control is complicated. The literature [14] proposes a high-boost converter, which has the advantage of simple structure. The literature [15] presents an interleaved boost converter characterized by an ultra-high boost ratio. The literature [16] proposes also an interleaved high-boost converter, which is characterized by inductive coupling and voltage superposition to realize a high-boost function.

All the proposed structures have higher voltage conversion ratios than conventional boost converters, but each has their own drawbacks. In the literature [4,5,6,7,9], although the structures are simple and easy to implement, the voltage conversion ratios are still not high enough, and the reason is that the voltage conversion ratios presented by references [4,7,9] have the same denominator of (1 − D), where D is the duty cycle and locates between 0 and 1. But, the numerators of [4,7,9] are 1 + D, 3, and 1 + D + n, respectively, where n is the turns ratio and generally locates between 1 and 3 to reduce leakage inductances. In the literature [4,6,7,8,9,10], the converters have floating switches, which require the use of additional isolation drivers, leading to an increase in cost. In the literature [9,10,11,12], the design is not easy, and the cost is too high due to the use of too many components. In the literature [4,8,9,11,12,13,14,16], the converters use the turns ratios of the coupled inductors to increase the voltage conversion ratios; however, there is a leakage problem of the coupled inductor, which is easy to resonate with the parasitic capacitance of the switch or diode, and hence, it is easy to cause a too large voltage spike on the switch or diode. Consequently, it is necessary to select components with higher voltage ratings or add additional buffers to overcome this drawback, which makes the circuit difficult to realize. In the literature [7,16,17,18,19,20,21], multiple voltage-doubling circuits are used to achieve higher voltage conversion ratios, resulting in an increase in the number of diodes and capacitors.

The high-boost converter proposed in this paper is derived from a charge-pump capacitor-based dual-boost converter [22]. Figure 1 shows the circuit structure of the charge-pump capacitor-based dual-boost inductor converter, which consists of two inductors, two diodes, two energy-transferring capacitors, one switch, one output diode, one output capacitor, and one load resistor. The voltage conversion ratio of V_o/V_i = 2/(1 − D) is not high, so the purpose of the circuit proposed in this paper is to improve the voltage conversion ratio of this converter based on three coils wound together with the energy stored in the leakage inductance recycled.

2. Proposed High-Boost Converter

2.1. Topology Description

Figure 2 shows the proposed high-boost converter, which consists of one switch, S₁; five diodes, D₁, D₂, D₃, D₄, and D₅; one energy-transferring capacitor, C_e; three stabilizing capacitors, C₁, C₂, and C₃; and one coupled inductor consisting of one primary-side winding N₁, one secondary-side winding N₂, one tertiary-side winding N₃, and one magnetizing inductor L_m. As for the load, it is represented by the output resistance R_o.

2.2. Operating Principles

Prior to analyzing the circuit operation, the relevant symbols and their assumptions are briefly explained: (i) the input voltage and current are denoted by V_i and i_i, respectively; (ii) the output voltage is signified by V_o; (iii) the capacitance of the energy-transferring capacitor C_e and voltage-stabilizing capacitors C₁, C₂, and C₃ are large enough so that the voltages across them are maintained at some certain values; (iv) the switching period is indicated by T_s; (v) the turns ratio n is defined to be N₃/N₁ = N₃/N₂; (vi) all the switches, diodes, and capacitors are regarded as ideal components; and (vii) the circuit is operated in the continuous conduction mode (CCM), and there are ten operating states over one switching period, as in Figure 3.

On the other hand, the voltages on and currents in components are defined as follows: (i) the gate driving signal for and voltage across on S₁ are represented by v_gs1 and v_ds1, respectively; (ii) the voltages on D₁, D₂, D₃, D₄, and D₅ are expressed by v_D1, v_D2, v_D3, v_D4, and v_D5, respectively; (iii) the voltages across C₁, C₂, C₃, and C_e are signified by V_C1, V_C2, V_C3, and V_Ce, respectively; (iv) the voltages on N₁, N₂, N₃, and L_m are denoted by v_N1, v_N2, v_N3, and v_N3, respectively; (v) the currents in N₁, N₂, N₃, L_m, and L_k3 are indicated by i_N1, i_N2, i_N3, i_Lm, and i_Lk3, respectively; and (vi) the currents in C₁, C₂, C₃, and C_e are i_C1, i_C2, i_C3, and i_Ce, respectively.

2.2.1. State 1: [$t_0\le t\le t_1$]

As shown in Figure 3 and Figure 4, the switch S₁ is on, the diodes D₁, D₂, and D₅ are on, and the diodes D₃ and D₄ are off. At the same time, the input voltage V_i charges the energy-transferring capacitor C_e, the energy stored in the leakage inductor L_k3 continues to charge the stabilizing capacitor C₃, and the stabilizing capacitors C₁ and C₂ provide energy to the load. At t = t₁, the magnetizing inductor L_m changes from demagnetization status to magnetization status. Once the stabilizing capacitor C₃ stops charging, state 1 comes to an end.

2.2.2. State 2: [$t_1\le t\le t_2$]

As shown in Figure 3 and Figure 5, the switch S₁ is still on, the diodes D₁, D₂, and D₅ are still on, and the diodes D₃ and D₄ are off. At the same time, the voltage across the magnetizing inductor L_m, called v_N3, equal to nV_i, is the mapped voltage from the primary side to the tertiary side, so the magnetizing inductor L_m is excited, the energy-transferring capacitor C_e is still in the state of charging, and the energy required by the load is supplied by the leakage inductor L_k3, the voltage-stabilizing capacitors C₁, C₂, and C₃, and the load. At t = t₂, the leakage inductor L_k3 has finished releasing energy, proceeding to state 2.

2.2.3. State 3: [$t_2\le t\le t_3$]

As shown in Figure 3 and Figure 6, the switch S₁ is still on, the diodes D₁, D₂, and D₄ are on, and the diodes D₃ and D₅ are cut off. At the same time, the magnetizing inductor L_m is still in the state of excitation and the energy required by the load is supplied by the voltage-stabilizing capacitors C₁, C₂, and C₃. Once the voltage-stabilizing capacitor C₂ stops supplying energy to the load, state 3 comes to an end.

2.2.4. State 4: [$t_3\le t\le t_4$]

As shown in Figure 3 and Figure 7, the switch S₁ is still on, the diodes D₁, D₂, and D₄ are still on, and the diodes D₃ and D₅ are still off. At the same time, the voltage-stabilizing capacitor C₂ is charging and the operating behavior of the other components is the same as that of state 3. Once the energy-transferring capacitor C_e stops charging, state 4 ends.

2.2.5. State 5: [$t_4\le t\le t_5$]

As shown in Figure 3 and Figure 8, the switch S₁ is cut off, the diodes D₃ and D₄ are conducted, and the diodes D₁, D₂, and D₅ are cut off. At the same time, the voltage across the magnetizing inductor L_m, called v_N3, equal to $0.5n(V_i+V_{Ce}-V_{C1})$, and hence, L_m is demagnetized, the energy-transfer capacitor C_e discharges and charges the voltage-stabilizing capacitor C₁, the energy stored in the leakage inductors L_k1, L_k2, and L_k3 release the energy to the load, and the energy required by the load is supplied by the input voltage V_i, the energy-transferring capacitor C_e, the magnetizing inductor L_m, and the voltage-stabilizing capacitor C₃. At t = t₅, the voltage-stabilizing capacitor C₂ stops charging, proceeding to state 5.

2.2.6. State 6: [$t_5\le t\le t_6$]

As shown in Figure 3 and Figure 9, the switch S₁ is still off, the diodes D₃ and D₄ are conducted, and the diodes D₁, D₂, and D₅ are cut off. The voltage-stabilizing capacitor C₂ releases energy to the load, and the operating behavior of the other components is the same as that of state 5. At t = t₆, state 6 ends.

2.2.7. State 7: [$t_6\le t\le t_7$]

As shown in Figure 3 and Figure 9, the operating behavior of the components is the same as that of state 6. Once the voltage-stabilizing capacitor C₃ stops supplying energy to the load, state 7 ends.

2.2.8. State 8: [$t_7\le t\le t_8$]

As shown in Figure 3 and Figure 10, the switch S₁ is still off, the diodes D₃ and D₅ are conducted, and the diodes D₁, D₂, and D₄ are cut off. At the same time, the magnetizing inductor L_m charges the voltage-stabilizing capacitor C₃, while the energy required for the load is supplied by the input voltage V_i, the energy-transfer capacitor C_e, the magnetizing inductor L_m, and the voltage-stabilizing capacitor C₂. The moment the energy-transferring capacitor C₁ stops charging, state 8 ends.

2.2.9. State 9: [$t_8\le t\le t_9$]

As shown in Figure 3 and Figure 11, the switch S₁ is cut off, the diodes D₃ and D₅ are conducted, and the diodes D₁, D₂, and D₄ are cut off. The magnetizing inductor L_m charges the voltage-stabilizing capacitor C₃. At the same time, the energy required by the load is supplied by the input voltage V_i, the energy-transferring capacitor C_e, the magnetizing inductor L_m, and the voltage-stabilizing capacitors C₁ and C₂. The moment the energy-transferring capacitor C_e stops supplying energy to the load, state 9 comes to an end.

2.2.10. State 10: [$t_9\le t\le t_0+T_s$]

As shown in Figure 3 and Figure 12, the switch S₁ is cut off, the diodes D₁, D₂, D₃, and D₄ are cut off, and the diode D₅ is still on. At the same time, the magnetizing inductor L_m charges the voltage-stabilizing capacitor C₃ and, together with the voltage-stabilizing capacitors C₁ and C₂, provides the energy required by the load. Once the switch S₁ is turned on, state 10 ends and returns to state 1, completing one cycle.

In order to simplify the analysis, the leakage inductances L_k1, L_k2, and L_k3 are not taken into consideration, so states 1, 2, 5, and 6 can be ignored.

From Figure 6 and Figure 7, it can be seen that the corresponding equations in states 3 and 4 are as follows:

$$\left\{\begin{array}{l}{v}_{N3}=n{V}_i\\ V_{C2}=v_{N3}\end{array}\right.$$

(1)

2.3. Voltage Gain

From Figure 9, Figure 10, Figure 11 and Figure 12, the corresponding equations for states 7, 8, 9, and 10 are as follows:

$$\left\{\begin{array}{l}{v}_{N3}=\frac{n(V_i+V_{Ce}-V_{C1})}{2}\\ V_{C3}=-v_{N3}\end{array}\right.$$

(2)

where

$$V_{Ce}=V_i$$

(3)

Therefore, (2) can be rewritten as

$$\left\{\begin{array}{l}{v}_{N3}=\frac{n(2V_i-V_{C1})}{2}\\ V_{C3}=-v_{N3}\end{array}\right.$$

(4)

Since the voltage across the magnetizing inductor L_m must obey the volt–second balance in the steady state, the following can be obtained:

$$n{V}_{i}D+\frac{n(2V_i-V_{C1})(1-D)}{2}=0$$

(5)

After rearranging (5), the following can be obtained:

$$\frac{V_{C1}}{V_i}=\frac{2}{1-D}$$

(6)

The output voltage is superimposed by the voltages across three voltage-stabilizing capacitors C₁, C₂, and C₃, that is,

$$V_o=V_{C1}+V_{C2}+V_{C3}$$

(7)

By substituting (1), (4), and (6) into (7), the following can be obtained:

$$V_o=\frac{2}{1-D}{V}_i+n{V}_i-\frac{n(2V_i-V_{C1})}{2}$$

(8)

Substituting (6) into (8) yields

$$\frac{V_o}{V_i}=\frac{n+2}{1-D}$$

(9)

2.4. Boundary Curve between Operating Modes of L_m

Assume that the converter has no loss in power conversion, i.e., $P_i=P_o$. Accordingly, by using (9), the input current can be expressed as

$$I_i=\frac{2+n}{1-D}{I}_o$$

(10)

where

$$I_o=\frac{V_o}{R_o}$$

(11)

Therefore, (10) can be rewritten as

$$I_i=\frac{n+2}{1-D}\times \frac{V_o}{R_o}$$

(12)

The average value of the magnetizing inductor current i_Lm is

$$I_{Lm}=I_{N3}-I_{Lk3}$$

(13)

Since the capacitor is equivalent to an open circuit under DC analysis, (13) can be rewritten as

$$I_{N3}=I_{Lm}$$

(14)

The transformer characteristics $\Sigma NI=0$ can be used to obtain a relationship between I_N1, I_N2, and I_N3:

$$I_{N1}+I_{N2}=n{I}_{N3}$$

(15)

By substituting (14) into (15), the following can be obtained:

$$n{I}_{Lm}=I_{N1}+I_{N2}$$

(16)

Since turns N₁ and N₂ are identical, that is, $I_{N1}=I_{N2}=I_N$, (16) can be rewritten as

$$I_{Lm}=\frac{2}{n}{I}_N$$

(17)

The average value of input current i_i is

$$I_i=I_{N1}+I_{N2}+I_{Ce}$$

(18)

According to the ampere–second balance, the average value of the current flowing through the capacitor C_e, called I_Ce, can be regarded as zero, so (18) can be rewritten as

$$I_i=2I_N$$

(19)

By substituting Equation (19) into (17), the following can be obtained:

$$I_{Lm}=\frac{1}{n}{I}_i$$

(20)

Finally, substituting (12) into (20) yields

$$I_{Lm}=\frac{n+2}{n(1-D)}\times \frac{V_o}{R_o}$$

(21)

The current ripple $\Delta i_{Lm}$ flowing through the magnetizing inductor L_m can be expressed as

$$\Delta i_{Lm}=\frac{v_{Lm}\Delta t}{L_m}=\frac{n{V}_{i}D{T}_s}{L_m}$$

(22)

Therefore, as $2I_{Lm}\ge \Delta i_{Lm}$, the magnetizing inductor L_m will be operated in the continuous conduction mode (CCM):

$$\begin{array}{l}2I_{Lm}\ge \Delta i_{Lm}\\ \Rightarrow \frac{2\times (n+2)}{n(1-D)}\frac{V_o}{R_o}\ge \frac{n{V}_{i}D{T}_s}{L_m}\\ \Rightarrow \frac{2L_m}{R_o{T}_s}\ge \frac{n^2{(1-D)}^{2}D}{{(n+2)}^2}\\ \Rightarrow K\ge K_{crit}(D)\end{array}$$

(23)

where $K=\frac{2L_m}{R_o{T}_s}$ and $K_{crit}(D)=\frac{n^2{(1-D)}^{2}D}{{(n+2)}^2}$.

From (23), if $K\ge K_{crit}(D)$, then the magnetizing inductor L_m operates in CCM; otherwise, it operates in the discontinuous current mode (DCM). Therefore, the boundary curve between CCM and DCM can be drawn as shown in Figure 13.

3. System and Component Specifications

Figure 14 shows the system configuration of the proposed high-boost converter, which consists of the main power stage and the negative-feedback control circuit. The main power stage is built up by a charge-pump capacitor-based dual-boost converter with three coils wound together. The feedback control circuit is composed of a voltage divider to obtain an analog signal of the output voltage, which is converted into a digital signal by an analog-to-digital converter (ADC), and then passed to an FPGA for computation to obtain the corresponding gate driving signal for the switch S₁ after a low-side gate driver.

The system and component specifications are as follows: (i) the input voltage is 12 V; (ii) the output voltage is 120 V; (iii) the rated output current is 0.5 A and the minimum output current is 0.05 A; (iv) the switching frequency is 100 kHz; (v) the product name of the switch S₁ is STP120NF10; (vi) the product name of the diodes D₁ and D₂ is MBRH2060CT; (vii) the product name of the diodes D₃, D₄, and D₅ is V20120C; (viii) the product name of the energy-transferring capacitor C_e is Rubycon with 470 µF; (ix) the product name of the voltage-stabilizing capacitors C₁, C₂, and C₃ is Rubycon with 100 µF; and (x) the product name of the FPGA is EP1C3T100.

4. Design Considerations

Table 1. Component specifications.

Component	Specification
Switch S1	STP120NF10
Diodes D1, D2	MBRH2060CT
Diodes D3, D4, D5	V20120C
Energy-Transferring Capacitor Ce	Ce = 470 µF/50 V Rubycon
Output Capacitors C1, C2, C3	C1 = 100 µF/100 V Rubycon C2 = C3 = 50 µF/100 V Rubycon
Coupled Inductor	PC44PQ26/25Z-12 Lm = 300 µH N1 = N2 = 11 turns, N3 = 22 turns Lk1 = Lk2 = 180 nH, Lk3 = 90 nH
Gate Driver	TC4420
FPGA	Cyclone I

shows the specifications of the proposed converter.

Since the general step-up-converter conversion ratio can theoretically be infinitely large but the elements have parasitic components, resulting in duty-cycle restrictions, the maximum of only operating the duty cycle is generally set at 0.6~0.7, which will be used in the design of the coupled inductor. Therefore, some system specifications can be substituted into (24) to obtain the turns ratio n:

$$\begin{array}{l}0.6\le \frac{V_o-(n+2)V_i}{V_o}\le 0.7\\ \Rightarrow 0.6\le \frac{120-(n+2)\times 12}{120}\le 0.7\\ \Rightarrow 1\le n\le 2\end{array}$$

(24)

In order to facilitate winding the coupled inductor, the turns ratio is taken as an integer:

$$n=2$$

(25)

By substituting (25) and some system specifications into (9), the corresponding duty cycle D can be obtained as

$$D=\frac{V_o-(n+2)V_i}{V_o}=\frac{120-(2+2)\times 12}{120}=0.6$$

(26)

The design of the coupled inductor is based on substituting the results of (25) and (26) into (23), and the magnetizing inductance of 300 μH is selected with the core named PC44PQ26/25Z-12.

5. Simulated Waveforms

The open-loop simulation of the proposed converter, under the condition of the rated output power of 60 W, was carried out to verify the feasibility of the proposed converter by using the obtained parameters shown in Table 1 and the PSIM 9.0 software. Figure 15 shows the simulated circuit of the proposed converter. Figure 16, Figure 17, Figure 18, Figure 19 and Figure 20 show the corresponding simulated waveforms.

As shown in Figure 19, the voltage on the energy-transferring capacitor C_e is stable at a certain value, and during the switch S₁ cut-off instant, the energy stored in the leakage inductance releases the energy to the load and charges the voltage-stabilizing capacitor C₁, so the current of the energy-transferring capacitor C_e, called i_Ce, will have a current surge. As shown in Figure 20, the voltage-stabilizing capacitors C₁, C₂, and C₃ are stable at some certain values. Therefore, the feasibility of the proposed structure is verified.

6. Measured Waveforms

The closed-loop experiment of the proposed converter, under the conditions of the minimum, half, and rated, was carried out to demonstrate the effectiveness of the proposed converter.

6.1. Steady-State Waveforms

Figure 21, Figure 22, Figure 23, Figure 24, Figure 25, Figure 26, Figure 27, Figure 28, Figure 29, Figure 30, Figure 31, Figure 32, Figure 33, Figure 34 and Figure 35 show the experimental waveforms to demonstrate the effectiveness of the proposed converter.

From Figure 21, Figure 22, Figure 26, Figure 27, Figure 31 and Figure 32, it can be seen that the voltage surge of the switch S₁ and the voltage surges of the diodes D₁, D₂, D₃, D₄ and D₅ will increase as the load increases, where the ringing phenomena of the diodes D₁ and D₂ are caused by the resonances of the parasitic capacitances of the diodes D₁ and D₂ with the leakage inductances of the coupled inductor and the parasitic inductance of the line when the switch S₁ is turned on. However, the simulation does not consider these two parasitic capacitances, so this phenomenon does not happen.

From Figure 24, Figure 25, Figure 29, Figure 30, Figure 34 and Figure 35, it can be seen that the voltage on the energy-transferring capacitor Ce, called V_Ce, and the voltage on the voltage-stabilizing capacitors C₁, C₂, and C₃, called V_C1, V_C2, and V_C3, respectively, are all approximately stable at certain values under any load. From Figure 23, Figure 28 and Figure 33, it can be seen that the current in the primary winding i_N1, the current in the secondary winding i_N2, and the current in the leakage inductance L_k3, called i_Lk3, will increase with the increase in the load current. As the switch S₁ is turned on, the part of the slope of the leakage inductance current i_Lk3 will be negative at light load, which is different from the negative slope of the current at half load and the positive slope of the current at rated load. The reason is that the voltage-stabilizing capacitor C₂ generates a larger current during charging, which results in a smaller slope of the current in the primary winding i_N1 and the secondary winding i_N2. Since $i_{N1}+i_{N2}=n{i}_{N3}$ and $i_{Lk3}=i_{N3}-i_{Lm}$ and the slope of the magnetizing current i_Lm does not change with the load, the slope of the leakage inductance current i_Lk3 at the light load and the medium load will be in the negative form. From Figure 26, Figure 29 and Figure 34, it can be seen that at the turn-on moment of the switch S₁, the input voltage V_i charges the energy-transferring capacitor C_e, and at the cut-off moment of the switch S₁, the energy stored in the leakage inductance releases the energy to the load and charges the regulator capacitor C₁, so the energy-transferring capacitor current i_Ce will generate current spikes at both the on and off moments of the switch S₁.

In summary, from the measured steady-state waveforms, it can be seen that the behavior of the proposed structure is consistent with the results of the analysis and simulation, so the effectiveness of the proposed structure is verified.

6.2. Dynamic Waveforms

Figure 36, Figure 37, Figure 38, Figure 39, Figure 40 and Figure 41 show the load transient responses.

Figure 36 and Figure 37 show the voltage responses due to the upload change from light load to half load and download change from half load to light load, respectively. As can be seen from Figure 36 and Figure 37, the corresponding output-voltage variations are about 6% and the corresponding recovery time is about between 50 ms and 60 ms. Figure 38 and Figure 39 show the output-voltage responses due to the upload change from half load to rated load and download change from rated load to half load, respectively. As can be seen from Figure 38 and Figure 39, the corresponding output-voltage variations are about 5% and the corresponding recovery time is about 50 ms. Figure 40 and Figure 41 show the output-voltage responses due to the upload change from light load to rated load and download change from rated load to light load, respectively. As can be seen from Figure 40 and Figure 41, the corresponding output-voltage variations are about 12% and the corresponding recovery time is about 65 ms. From these figures, it can be seen that the more the load change is, the more voltage variation and the larger the recovery time.

6.3. Efficiency Measurement

Figure 42 shows the block diagram of efficiency measurement. A shunt resistor is connected in series with the input-current path, and a digital meter (Fluke 8050A) is used to measure the voltage on this resistor to obtain the input-current value, and a digital meter (Fluke 179) is used to measure the input voltage to obtain the input power. On the output side, an electronic load (Prodigit 3314F) is used to provide the load current required by the converter, and the output voltage is measured using a digital meter (Fluke 179) to obtain the output power. Eventually, the resulting input and output powers are used to calculate the efficiency of the actual circuit operation. Figure 43 shows the curve of efficiency versus load current.

From Figure 43, it can be seen that the efficiency under any load is above 89.4%, and the maximum efficiency is 92.61%.

6.4. Experimental Setup

Figure 44 displays a photo of the experimental setup.

7. Circuit Comparison

Table 2. Comparison of the existing and proposed converters.

Reference Number	CCM Voltage Gain	Total Component No.	Diode No.	Switch No.	Inductor No.	Capacitor No.	Switching Frequency	Efficiency
[18]	$\frac{2+\left(2-D\right)n_{21}-(1+D)n_{31}}{\left(1-D\right)(1-n_{31})}$	14	4	1	4	1	55 kHz	94.40%
[19]	$\frac{2+\left(2-D\right)n_{21}-(1+D)n_{31}}{\left(1-D\right)(1-n_{31})}$	16	5	1	4	6	60 kHz	95.10%
[20]	$\frac{2{(1-D)}^2}{1-4D+{2D}^2}$	16	6	1	4	5	50 kHz	94.70%
[21]	$\frac{(2+D)}{{(1-D)}^2}$	16	5	2	3	6	33 kHz	95.22%
[22]	$\frac{(2+D)}{{(1-D)}^2}$	8	3	1	2	2	100 kHz	92.50%
Proposed	$\frac{n+2}{1-D}$	13	5	1	1	4	100 kHz	92.61%

displays a comparison of the existing and proposed converters. In this table, the comparison items are in terms of CCM voltage gain, total number of components, number of diodes, number of inductors, switching frequency, and efficiency. The proposed converter has a relatively small number of components and a relatively linear voltage gain. But, because the switching frequency of the proposed converter is relatively high, the corresponding efficiency is relatively low.

8. Conclusions

In this paper, a new high-boost converter is proposed, whose main structure is improved based on the converter with a charge-pump capacitor and dual-boost inductor. This structure is made by adding an additional voltage doubler and a set of coils to the existing circuit to increase the voltage conversion ratio, and three coils are wound on the same magnetic core, reducing the overall circuit size. In addition, the energy stored in the leakage inductance of the coupled inductor can be recovered. Accordingly, the efficiency can be up to 92.61%.