A UV-C LED Sterilization Lamp Driver Circuit with Boundary Conduction Mode Control Power Factor Correction ^†

Abstract

The increasing prevalence of common cold viruses and bacteria in daily life has heightened interest in sterilization lamp technologies. Compared with traditional mercury-based ultraviolet (UV) lamps, modern UV lamps offer advantages including extended operational lifespan, high energy efficiency, compact form factor, and the absence of hazardous materials, rendering them both safer and environmentally sustainable. In particular, UV-C LED lamps, which emit at short wavelengths, are capable of disrupting the molecular structure of DNA or RNA in microbial cells, thereby inhibiting cellular replication and achieving effective disinfection and sterilization. Conventional UV-C LED sterilization lamp driver circuits frequently employ a two-stage architecture, which requires a large number of components, occupies substantial physical space, and exhibits reduced efficiency due to multiple stages of power conversion. To address these limitations, this paper proposes a UV-C LED sterilization lamp driver circuit for an AC voltage supply, employing boundary conduction mode (BCM) control with integrated power factor correction (PFC). The proposed single-stage, single-switch topology combines a buck PFC converter and a flyback converter while recovering transformer leakage energy to further improve efficiency. Compared with conventional two-stage designs, the proposed circuit reduces the number of power switches and components, thereby lowering manufacturing cost and enhancing overall energy conversion efficiency. The operating principles of the proposed driver circuit are analyzed, and a prototype is developed for a 110 V AC input with an output specification of 10.8 W (90 V/0.12 A). Experimental results demonstrate that the prototype achieves an efficiency exceeding 92%, a power factor of 0.91, an output voltage ripple of 1.298%, and an output current ripple of 4.44%.

1. Introduction

Microorganisms, including bacteria, viruses, cysts, and molds, utilize germination or spore formation as adaptive reproductive mechanisms to ensure persistence and survival under adverse environmental conditions. When subjected to irradiation from sterilization lamps—particularly those emitting germicidal wavelengths—the incident photons penetrate microbial cell walls and membranes, subsequently inducing irreversible damage to nucleic acids through the disruption of DNA molecular structures. This photochemical degradation compromises the integrity of the genetic material, thereby inhibiting replication processes or leading to complete cell lysis, ultimately achieving microbial inactivation. Sterilization lamps have therefore emerged as an effective and non-chemical intervention for disinfection and sterilization. Their deployment spans a wide spectrum of sectors, including but not limited to healthcare facilities, food and beverage processing, public transportation systems, commercial and office environments, residential settings, hospitality industries, educational institutions, retail and grocery operations, water treatment plants, and agricultural production. The breadth of these applications underscores the operational versatility, broad-spectrum antimicrobial efficacy, and pivotal role of sterilization lamps in safeguarding public health and ensuring biosafety across diverse environmental contexts [1,2,3,4,5].

Ultraviolet (UV) mercury lamps have long been utilized as conventional sterilization light sources due to their relatively low cost and high luminous output, making them suitable for large-scale applications in environments such as hospitals, factories, restaurants, and schools. Despite these advantages, UV mercury lamps exhibit notable limitations, including prolonged warm-up times, relatively short operational lifespans, and the presence of mercury, which poses environmental and regulatory concerns. Ultraviolet radiation, which is invisible to the human eye, spans wavelengths from 100 to 400 nm and can be classified into three spectral regions: long-wave UV-A (320–400 nm), medium-wave UV-B (280–320 nm), and short-wave UV-C (100–280 nm). The UV-C band—also referred to as deep ultraviolet—possesses potent germicidal properties. Its sterilization mechanism involves the absorption of short-wave UV photons by microbial nucleic acids (DNA or RNA), leading to the formation of photoproducts that disrupt molecular structures, inhibit cell replication, and ultimately inactivate the microorganism [6,7,8].

In recent years, UV-C light-emitting diodes (LEDs) have emerged as an alternative to traditional mercury-based UV sources, offering several advantages, including environmental sustainability, energy efficiency, extended operational lifespan, instant-on capability without preheating, mercury-free composition, and the potential for compact and portable designs. These features have facilitated the increasing adoption of UV-C LEDs in applications such as air and surface disinfection within residential, educational, healthcare, commercial, and retail settings. However, the germicidal performance of UV-C LEDs is influenced by their inherently low penetration depth, which makes the radiation susceptible to attenuation by airborne particulates or surface dust. Incomplete nucleic acid damage under suboptimal irradiation conditions may permit microbial DNA repair mechanisms, potentially restoring viability. Therefore, achieving effective disinfection requires careful optimization of the irradiation parameters, including source-to-target distance, exposure duration, and radiant intensity, to ensure complete microbial inactivation [9,10,11,12,13].

Table 1. Comparison between traditional UV mercury lamps and UV-C LEDs.

Parameter	UV Mercury Lamp	UV-C LED
Peak wavelength	~254 nm	255–280 nm (tunable)
Preheating requirement	Yes	No
Warm-up time	1–5 min	Instant-on
Lamp life	5000–8000 h	10,000–50,000 h
Environmental impact	Contains mercury (toxic, disposal issues)	Mercury-free (environmentally friendly)
Size & form factor	Large, fragile glass tube	Compact, robust, solid-state
Initial cost	Low	Higher
Radiant power output	High (suitable for large-area applications)	Lower (requires optimized placement)
Energy efficiency	Moderate	High

summarizes the key performance differences between traditional UV mercury lamps and UV-C LEDs. As shown, UV-C LEDs offer distinct advantages in terms of environmental safety, lifetime, and operational flexibility, whereas mercury lamps retain an advantage in initial cost and radiant power output. This comparison underscores the technological shift toward UV-C LED-based disinfection and sterilization systems in both commercial and domestic applications.

In conventional off-line power converters, the front-end stage generally comprises a full-wave rectifier bridge followed by a filtering capacitor, which converts the AC mains voltage into an unregulated DC bus. The filtering capacitor must possess sufficiently large capacitance to ensure that the ripple voltage superimposed on the DC output remains minimal. As a result, the instantaneous AC line voltage is often lower than the voltage across the filtering capacitor, causing the rectifier diodes to conduct only during a brief portion of each half-cycle. Consequently, the input current drawn from the AC mains manifests as a sequence of narrow pulses with amplitudes typically 5 to 10 times higher than the average DC current. This pulsed current profile leads to several adverse effects, including significantly elevated peak and RMS input currents, increased distortion of the AC line voltage, and inefficient utilization of the power system’s energy capacity. To address these issues, a power factor correction (PFC) pre-regulator is often inserted between the rectifier bridge and the filtering capacitor. By employing high-frequency switching techniques, the PFC circuit enables the converter to draw a quasi-sinusoidal input current that is nearly in phase with the AC line voltage. This approach significantly improves the power factor and enhances the overall efficiency of energy transfer from the mains while mitigating the aforementioned drawbacks associated with conventional front-end rectification.

Conventional UV-C LED sterilization lamp driver circuits incorporating PFC typically adopt a two-stage topology. In this configuration, the first stage comprises an AC–DC power converter with integrated PFC functionality, while the second stage consists of a DC–DC converter that supplies the regulated DC power required by the UV-C LED sterilization lamp [13]. Such two-stage driver circuits necessitate independent control units for both the front-end and back-end stages, thereby increasing circuit complexity and component count. Furthermore, the overall efficiency of these designs is inherently constrained by the sequential two-stage energy conversion process. To mitigate these drawbacks—specifically, to reduce the number of power switches and associated circuit components while enhancing overall conversion efficiency—various studies [14,15,16,17,18,19,20,21,22] have proposed single-stage, single-switch LED driver architectures. These designs integrate the AC–DC power conversion and DC–DC regulation functions into a unified topology, rendering them suitable for UV-C LED sterilization lamp applications.

The buck-type PFC converter represents a viable topology for regulating direct current (DC) output voltage while simultaneously improving the power factor in alternating current to direct current (AC–DC) conversion stages. Unlike the widely adopted boost PFC converter—which inherently steps up the voltage and necessitates that the output voltage exceed the peak amplitude of the AC input—the buck PFC converter operates in a step-down mode, enabling the output voltage to remain below the peak value of the input AC voltage. This operational characteristic renders the buck PFC topology particularly advantageous for low-voltage applications, such as light-emitting diode (LED) drivers and battery charging systems, where the downstream load operates at relatively modest voltage levels.

A key advantage of the buck PFC converter lies in its reduced voltage stress on both the power switching device and the output diode, resulting in lower conduction and switching losses compared to its boost counterpart, especially under low output voltage conditions. In addition, the topology inherently provides continuous inductor current and a low voltage conversion ratio, both of which contribute to diminished output voltage ripple. The voltage stress imposed on the switching element typically remains below the peak value of the input AC voltage, while the reverse voltage stress on the diode is likewise minimized. These characteristics facilitate the use of lower-voltage-rated and more cost-effective components.

Despite these merits, buck-type PFC converters are subject to inherent limitations. Most notably, they are incapable of producing an output voltage higher than the peak of the AC input, thereby restricting their applicability in certain high-voltage scenarios. Moreover, near the zero-crossing points of the AC input waveform, the input current tends to become discontinuous, leading to increased harmonic distortion and deterioration in power quality. This behavior further contributes to a degradation in power factor, particularly in the vicinity of the zero-crossing intervals. To mitigate these issues, advanced control techniques—such as boundary conduction mode (BCM) combined with input current shaping strategies—have been proposed and investigated in the literature [23,24,25,26,27].

The flyback converter, which employs a single switching device, a transformer, and a small number of passive components, remains one of the most widely adopted power conversion topologies, particularly for low-to-medium power AC–DC and DC–DC applications. Its simplicity and ability to provide galvanic isolation with multiple output voltages make it attractive for cost-sensitive designs. However, the inherent leakage inductance of the flyback transformer can result in elevated peak currents, which, in turn, increase switching and conduction losses. Consequently, the inclusion of additional snubber or clamp circuitry is generally recommended to suppress voltage spikes and improve overall reliability.

Building upon and extending the concepts presented in [28,29], this paper proposes a single-switch UV-C LED sterilization lamp driver circuit. The main topology combines a buck PFC converter and a flyback converter into a single-stage, single-switch buck–flyback AC–DC power converter with input-current shaping capability by using a boundary conduction mode control scheme. Moreover, the proposed driver circuit incorporates an energy recovery mechanism for the leakage inductance of the transformer, thereby improving overall efficiency. In addition, the proposed design employs a boundary conduction mode (BCM) control strategy and builds upon the foundational concepts presented in [28,29], which utilize a discontinuous conduction mode (DCM) control scheme while adopting the same main circuit topology to achieve PFC. In contrast to the approaches in [28,29], the proposed work increases the output power by delivering a higher output current at the same output voltage. Additionally, the circuit efficiency is enhanced, demonstrating improved overall performance.

The structure of this paper is outlined as follows: Section 2 presents an analysis of the operational modes and elaborates on the control principles of the proposed UV-C LED sterilization lamp driver circuit, which incorporates boundary conduction mode power factor correction. Section 3 details the design methodology for key circuit parameters of the proposed driver. Section 4 reports the experimental results obtained from a prototype implementation of the driver circuit for UV-C LED sterilization applications. Finally, Section 5 summarizes the main findings and provides concluding remarks.

2. Analysis of Operational Modes and Elaboration on the Control Concepts in the Proposed UV-C LED Sterilization Lamp Driver Circuit with Boundary Conduction Mode Power Factor Correction

Figure 1 illustrates the proposed UV-C LED sterilization lamp driver circuit, which operates in BCM with integrated PFC. The topology is implemented as a single-stage, single-switch buck–flyback AC–DC converter, combining the functional characteristics of a buck PFC converter and a flyback converter into a unified architecture. The buck PFC subcircuit comprises a filter inductor L_F, a filter capacitor C_F, a full-wave rectifier (consisting of D₁, D₂, D₃, and D₄), a capacitor C_REC, a power switch S₁, a diode D₅, a transformer T_R, and an output capacitor C₂. The flyback converter subcircuit shares certain components with the buck stage, including C_REC, S₁, and D₅, and further incorporates an additional diode D₆, the transformer T_R with N_P turns on the primary winding and N_S turns on the secondary winding, as well as two output capacitors, C₁ and C₂. The output of the converter drives the UV-C LED sterilization lamp, enabling efficient AC–DC conversion with inherent power factor correction.

Figure 2 is the equivalent circuit diagram of the UV-C LED sterilization lamp driver circuit with BCM PFC in the operational mode analysis, and Figure 3 shows the theoretical waveforms of the components of the UV-C LED sterilization lamp driver circuit with BCM PFC. In this study, the analysis of the operating modes of the proposed UV-C LED sterilization lamp driver circuit, which incorporates a BCM PFC scheme, is conducted under the following assumptions to facilitate analytical tractability:

(a)

The AC input voltage v_AC is assumed to be rectified by a full-wave bridge and subsequently filtered by capacitor C_REC, resulting in a DC voltage source denoted as V_REC.

(b)

To achieve input current shaping, the power switch is operated under a transition mode pulse-width modulation (PWM) strategy, characterized by a fixed turn-on time and variable switching frequency. Consequently, the duty cycle of the power switch S₁ is set to 0.5. For simplicity, the effects of the switch’s parasitic diode and parasitic capacitance are neglected.

(c)

The magnetizing inductor L_M is designed to operate in boundary conduction mode with a peak current control scheme to ensure effective power factor correction.

(d)

The leakage inductances on the primary and secondary sides of the transformer T_R are represented by L_lk1 and L_lk2, respectively.

(e)

All remaining circuit components are considered ideal for the purpose of this analysis.

Figure 2. Equivalent circuit of the proposed UV-C sterilization lamp driver circuit with BCM PFC while analyzing the operational modes.

Figure 2. Equivalent circuit of the proposed UV-C sterilization lamp driver circuit with BCM PFC while analyzing the operational modes.

Figure 3. Theoretical waveforms of the UV-C sterilization lamp driver circuit with BCM PFC.

Figure 3. Theoretical waveforms of the UV-C sterilization lamp driver circuit with BCM PFC.

Operation Mode 1 (t₀ ≤ t < t₁): Figure 4 shows the equivalent circuit of the UV-C sterilization lamp driver circuit with BCM PFC during Mode 1. At t₀, the switch S₁ conducts, and the diode D₅ is reverse-biased. The full-wave rectified voltage source V_REC supplies energy to the magnetizing inductance L_M and the primary leakage inductance L_lk1 as well as the output capacitor C₂ via the power switch S₁, and the magnetizing inductor current I_LM and the inductor current I_Llk1 of primary-side leakage inductance in the transformer rise linearly. Output capacitors C₁ and C₂ supply energy to the load. Operation Mode 1 ends when power switch S₁ is turned off.

Operational Mode 2 (t₁ ≤ t < t₂): Figure 5 shows the equivalent circuit of the UV-C sterilization lamp driver circuit with BCM PFC during Mode 2. After the power switch of the equivalent circuit is cut off at t₁, the diode D₅ is forward-biased, the primary side leakage inductor L_lk1 provides energy to the output capacitor C₂ through the diode D₅, and the transformer primary side leakage current I_Llk1 decreases linearly. The magnetizing inductor L_M provides energy to the secondary side and through diode D₆ to the output capacitor C₁, the magnetizing inductor current I_LM decreases linearly, and the transformer secondary leakage current I_Llk2 increases linearly. Output capacitors C₁ and C₂ provide energy to the load. When the transformer primary leakage inductance current I_Llk1 equals zero, the operational Mode 2 ends.

Operational Mode 3 (t₂ ≤ t < t₃): Figure 6 shows the equivalent circuit of the UV-C sterilization lamp driver circuit with BCM PFC during Mode 3. In the equivalent circuit, at t₂, the power switch S₁ is still cut off. The magnetizing inductor L_M continues to supply energy to the secondary side and is supplied to the output capacitor C₁ via diode D₆. The magnetizing inductor current I_LM decreases linearly, and the transformer secondary leakage current I_Llk2 also decreases linearly. Output capacitors C₁ and C₂ continuously supply energy to the load. When the power switch S₁ is turned on again, operational Mode 3 ends and returns to operational Mode 1.

Figure 7 illustrates the conceptual control diagram of the proposed UV-C LED sterilization lamp driver circuit operating with BCM PFC. In the design of the driver and control circuitry, the magnetizing inductor L_M of the transformer T_R is configured to operate in BCM and is coordinated with an integrated circuit (IC) that incorporates a peak current control function. To achieve power factor correction, the control strategy involves maintaining a fixed on-time for the power switch S₁ while modulating its switching frequency. This approach enables the input current to closely track the shape of the input voltage waveform, thereby improving power factor performance.

Furthermore, the process by which the control signal v_GS is generated and output by the peak current control IC to drive the power switch S₁ is briefly outlined in the following discussion:

(1)

The output voltage of the UV-C LED sterilization lamp is obtained from the detection resistors R_v1 and R_v2, and then sent to the error amplifier through the output voltage detection and feedback circuit and feedback compensation circuit to compare with the 2.5-volt DC voltage level to obtain a voltage signal V_c. After the AC input voltage is rectified by a full-wave bridge rectifier, a voltage signal V_rec’ is obtained from the detection resistors R_rec1 and R_rec2, which, together with the voltage signal V_c, are fed into a multiplier to obtain a voltage signal V_m. The current signal i_DS’ flowing through the power switch S₁ is converted into a voltage signal V_s by the detection resistor R_s. A comparator compares the two voltage signals V_m and V_s and sends them to the Reset (R) pin of the Flip-Flop, and when the voltage signals V_m and V_s are equal, the power switch S₁ is turned off by this mechanism.

(2)

The current i_LM’ flowing through the magnetizing inductor L_M is detected by adding the third winding N_c of the transformer T_R and connecting a circuit R_c in series, and the return current i_LM’ is proportional to the output current of the UV-C LED sterilization lamp. This return current i_LM’ is proportional to the output current of the UV-C LED sterilization lamp. The return current i_LM’ is then fed to the Set (S) pin of the inverter through the zero current detector. Since the magnetizing inductor is designed to operate in the boundary conduction mode, when the current i_LM’ flowing through the magnetizing inductor L_M drops to zero, the power switch S₁ is activated by this mechanism.

3. Design Guideline of the UV-C LED Sterilization Lamp Driver Circuit with BCM PFC

3.1. Design Guideline of the Magnetizing Inductor L_M in the Transformer T_R

Figure 8 illustrates the theoretical waveforms of the magnetizing inductor current i_LM(t), its peak value i_LM-pk(t), the input utility-line current i_AC(t), and the utility-line voltage v_AC(t) under BCM operation with a peak current control scheme. These idealized waveforms are representative of conventional boost-type PFC converters, which are widely adopted due to their capability to maintain continuous input current with minimal distortion under grid-connected conditions. The peak value of input current, denoted as I_AC-pk, is given by

$$I_{AC-pk}={\displaystyle \frac{\sqrt{2}{P}_{IN}}{v_{AC-rms}}}={\displaystyle \frac{\sqrt{2}{P}_O}{\eta v_{AC-rms}}}$$

(1)

where v_AC-rms is the rms value of the input utility-line voltage; η is the estimated circuit efficiency. Besides, the output power P_O is calculated by multiplying the input power P_IN by the estimated circuit efficiency η. The peak value of the magnetic induction current I_LM-pk is twice the peak value of the input current I_AC-pk, which can be expressed as

$$I_{LM-pk}=2I_{AC-pk}={\displaystyle \frac{2\sqrt{2}{P}_O}{\eta v_{AC-rms}}}$$

(2)

With a η of 0.85, a V_O of 90 V, a P_O of 10.8 W, and a v_AC-rms of 110 V, the peak value of the magnetic induction current I_LM-pk is calculated by

$$I_{LM-pk}={\displaystyle \frac{2\sqrt{2}{P}_O}{\eta v_{AC-rms}}}={\displaystyle \frac{2\sqrt{2}\times 10.8}{0.85\times 110}}=0.327A$$

Referring to Figure 8, the difference in amplitude between peak and low levels of the magnetizing inductor current, indicated as ∆I, is obtained by the following equation:

$$\Delta I={\displaystyle \frac{\sqrt{2}{v}_{AC-rms}\left|\sin \left(2\pi f_{AC}t\right)\right|-\frac{1}{2}{V}_O}{L_M}}{T}_{ON}={\displaystyle \frac{V_O}{2L_M}}{T}_{OFF}$$

(3)

where T_ON and T_OFF are turn-on and turn-off times of the power switch, respectively.

In order to design the magnetic inductor under BCM, the turn-on time of the power switch can be expressed by

$$T_{ON}(2\pi f_{AC}t)={\displaystyle \frac{L_M{I}_{LM-pk}\left|\sin (2\pi f_{AC}t)\right|}{\sqrt{2}{v}_{AC-rms}\left|\sin (2\pi f_{AC}t)\right|-\frac{1}{2}{V}_O}}$$

(4)

While sin(2πf_ACt) is equal to unity, the maximum level of T_ON, indicated as T_ON-max, is given by

$$T_{ON-\max }={\displaystyle \frac{L_M{I}_{LM-pk}}{\sqrt{2}{v}_{AC-rms}}}$$

(5)

The turn-off time of the power switch can be expressed as

$$T_{OFF}(2\pi f_{AC}t)={\displaystyle \frac{2L_m{I}_{Lm-pk}\left|\sin (2\pi f_{AC}t)\right|}{V_O}}$$

(6)

The minimum switching frequency f_{SW min} of the power switch is shown in the following equation:

$$\begin{array}{l}{f}_{SW-\min }(2\pi f_{AC}t)={\displaystyle \frac{1}{T_{ON}(2\pi f_{AC}t)+T_{OFF}(2\pi f_{AC}t)}}\\ ={\displaystyle \frac{V_O\left(\sqrt{2}{v}_{AC-rms}\left|\sin \left(2\pi f_{AC}t\right)\right|-\frac{1}{2}{V}_O\right)}{L_M{I}_{LM-pk}2\sqrt{2}{v}_{AC-rms}\left|\sin \left(2\pi f_{AC}t\right)\right|}}\end{array}$$

(7)

Rearranging (7), the magnetic inductor L_m operated in BCM can be expressed as

$$L_M={\displaystyle \frac{V_O\left(\sqrt{2}{v}_{AC-rms}\left|\sin \left(2\pi f_{AC}t\right)\right|-\frac{1}{2}{V}_O\right)}{I_{LM-pk}2\sqrt{2}{v}_{AC-rms}\left|\sin \left(2\pi f_{AC}t\right)\right|f_{SW-\min }}}$$

(8)

Referring to (6) with a V_O of 90 V, a P_O of 10.8 W, a peak value of the magnetic induction current I_LM-pk of 0.327 A, a minimum switching frequency f_{SW min} of 35 kHz, and a v_AC-rms of 110 V, the inductance of the magnetizing inductor L_M is calculated by

$$\begin{array}{l}{L}_M={\displaystyle \frac{V_O\left(\sqrt{2}{v}_{AC-rms}\left|\sin \left(2\pi f_{AC}t\right)\right|-\frac{1}{2}{V}_O\right)}{I_{LM-pk}2\sqrt{2}{v}_{AC-rms}\left|\sin \left(2\pi f_{AC}t\right)\right|f_{SW}}}\\ ={\displaystyle \frac{90\times \left(\sqrt{2}\times 110-\frac{1}{2}\times 90\right)}{0.327\times 2\sqrt{2}\times 110\times \mathrm{35,000}}}=2.79 \mathrm{mH}\end{array}$$

In addition, a magnetizing inductance L_M of 2.67 mH was wound to realize the prototype driver circuit.

3.2. Design Guideline of the Turns Ratio of Transformer T_R

According to the volt-second balance principle, the voltage across the magnetizing inductor L_M during the switch-on interval of S₁ multiplied by the corresponding on-time is equal to the voltage across L_M during the switch-off interval multiplied by the corresponding off-time. This relationship can be expressed mathematically as follows:

$$\left(\sqrt{2}{v}_{AC-rms}-{\displaystyle \frac{V_O}{2}}\right)Duty{T}_S={\displaystyle \frac{N_P}{N_S}}{\displaystyle \frac{V_O}{2}}\left(1-Duty\right)T_S$$

(9)

The definition of turns-ratio n of transformer T_R can be determined by dividing the primary-side winding N_P by the secondary-side winding N_S, and can be expressed as

$$n={\displaystyle \frac{N_P}{N_S}}={\displaystyle \frac{\left(2\sqrt{2}{v}_{AC-rms}-V_O\right)Duty}{V_O\left(1-Duty\right)}}$$

(10)

Referring to (10) with v_AC-rms of 110 V, a Duty of 0.5, and a V_O of 90 V, turns-ratio n is calculated by

$$n={\displaystyle \frac{N_P}{N_S}}={\displaystyle \frac{\left(2\sqrt{2}{v}_{AC-rms}-V_O\right)Duty}{V_O\left(1-Duty\right)}}={\displaystyle \frac{\left(2\sqrt{2}\times 110-90\right)\times 0.5}{90\times \left(1-0.5\right)}}=2.456$$

In addition, the turns-ratio n is selected to be 2 when implementing the prototype circuit.

3.3. Design Guideline of the Output Capacitors C₁ and C₂

The design of the output capacitance C_O, implemented through output capacitors C₁ and C₂, is governed by several parameters, including the nominal DC output voltage, the allowable overvoltage margin, the output power level, and the desired output voltage ripple. The voltage ripple, ∆V_O, is defined as half of the peak-to-peak fluctuation in output voltage occurring at twice the line frequency. This ripple is a function of the output capacitance C_O and the peak capacitor current, which is approximately equal to the output current I_O, and can be mathematically expressed as described in [30].

$$\Delta V_O=I_O\sqrt{{\left({\displaystyle \frac{1}{2\pi \times 2f_{AC}\times C_O}}\right)}^2+{R_{ESR}}^2}$$

(11)

By neglecting the equivalent series resistance (ESR), denoted as R_ESR, of the output capacitor, the required capacitance C_O for the output capacitors C₁ and C₂ can be formulated as

$$C_O\ge {\displaystyle \frac{I_O}{4\pi f_{AC}\times \Delta V_O}}={\displaystyle \frac{P_O}{4\pi f_{AC}\times V_O\times \Delta V_O}}$$

(12)

In practical design, the voltage ripple ∆V_O is generally selected to be 1% to 5% of the nominal output voltage.

For the case under consideration, substituting the circuit parameters into (12), with a P_O of 10.8 W, an f_AC of 60 Hz, a V_O of 90 V, a ΔV_O of 1.35 V (corresponding to 1.5% of V_O), the calculated value of capacitance C_O is given by

$$C_O\ge {\displaystyle \frac{P_O}{4\pi f_{AC}\times V_O\times \Delta V_O}}={\displaystyle \frac{10.8}{4\pi \times 60\times 90\times 1.35}}=117.89 \mathsf{\mu }\mathrm{F}$$

When implementing the prototype circuit, to suppress output voltage ripple below 1.35 V, the capacitance values of output capacitors C₁ and C₂ were selected as 150 μF, and the voltage rating of the capacitors was selected as 250 V.

4. Experimental Results of the Proposed UV-C LED Sterilization Lamp Driver Circuit with BCM PFC

Table 2. Specifications of the proposed UV-C sterilization lamp driver circuit incorporating BCM PFC.

Parameter	Value
Input AC Voltage vAC	110 V
Rated Output Power PO	10.8 W
Rated Output Voltage VO	90 V
Rated Output Current IO	120 mA

summarizes the specifications of the proposed UV-C LED sterilization lamp driver circuit incorporating BCM PFC, as developed and implemented in this study. The driver circuit was designed to operate with an RMS AC input voltage of 110 V. The output parameters were determined based on the specifications of the UV-C LED sterilization lamp, including an output power of 10.8 W, an output voltage of 90 V, and an output current of 120 mA.

Table 3. Key components employed in the implementation of the proposed UV-C sterilization lamp driver circuit with BCM PFC.

Component	Value
Filter Inductor LF	3 mH
Filter Capacitor CF	220 nF
Diodes D1, D2, D3, D4	MUR460
Capacitor CREC	470 nF
Power Switch SB	SPW47N60C3
Turns Ratio n of Transformer TR	n = NP/NS = 2
Magnetizing Inductor LM	2.67 mH
Leakage Inductors Llk1, Llk2	26.7 μH, 9.61 μH
Diodes D5, D6	MUR460
Output Capacitor CO1, CO2	150 μF

lists the key components employed in the implementation of the proposed UV-C LED sterilization lamp driver circuit.

Figure 9 illustrates the experimentally measured waveforms of the leakage inductor currents I_Llk1 and I_Llk2. Specifically, I_Llk1 flows when the primary-side power switch S₁ is turned on, whereas I_Llk2 conducts when S₁ is turned off. From the measured waveforms, it can be observed that once the secondary-side leakage current I_Llk2 decreases to zero, the primary-side leakage current I_Llk1 immediately begins to increase linearly. This behavior confirms that the magnetizing inductor L_M of the transformer operates in the BCM.

Figure 10 shows the measured waveforms of the output voltage V_O and output current I_O. The average values of output voltage and current were recorded as 90.87 V and 126 mA, respectively. We acknowledge that the measured output current in Figure 10 is presented over a relatively short time scale, which may limit the visibility of waveform characteristics. To address this, additional measurements of both the output voltage and output current were conducted over an extended time scale. This approach enables accurate determination of the peak-to-peak values of the output waveforms, which are essential for calculating the output voltage ripple factor and output current ripple factor of the proposed UV-C LED sterilization lamp driver circuit operating under BCM with PFC.

Table 4. Measurement of output voltage ripple factor and output current ripple factor of UV-C LED sterilization lamp driver circuit with BCM PFC.

Parameters	Values
Mean value of the output voltage	90.87 V
Peak-to-peak value of the output voltage	1.18 V
Ripple factor of the output voltage	1.298%
Mean value of the output current	126 mA
Peak-to-peak value of the output current	5.6 mA
Ripple factor of the output current	4.44%

presents the calculated output voltage and current ripple factors of the prototype UV-C LED sterilization lamp driver circuit operating with BCM PFC under a 110 V_RMS AC input. As shown in Table 4, the voltage ripple factor is determined by dividing the peak-to-peak voltage ripple of 1.18 V by the average output voltage of 90.87 V, resulting in a value of 1.298%. Similarly, the current ripple factor is obtained by dividing the peak-to-peak output current of 5.6 mA by the average current of 126 mA, yielding a ripple factor of 4.44%.

Figure 11 illustrates the experimentally measured waveforms of the AC input voltage (v_AC) and input current (i_AC). The observed waveforms confirm that the proposed driver circuit successfully achieves PFC functionality. Nevertheless, it is evident that the input current exhibits discontinuous behavior in the vicinity of the zero-crossing points of the AC input voltage. This discontinuity results in elevated current harmonic distortion and a consequent degradation in overall power quality. The deterioration in power factor is particularly pronounced during these zero-crossing intervals.

This behavior is intrinsically associated with the operational characteristics of the buck converter topology implemented in the proposed circuit architecture. Specifically, as the instantaneous input voltage approaches zero, the input current also tends to diminish to zero, thereby entering a discontinuous conduction mode. This leads to interruptions in current flow, contributing to the observed harmonic distortion.

Quantitative measurements obtained using a precision power analyzer (Tektronix PA4000) indicate a power factor of 0.9164 and a total harmonic distortion (THD) of the input current of 29.513%. These results further substantiate the impact of the buck-type topology on the input current waveform and power quality characteristics. This behavior highlights the key distinction between the input current profiles of boost-type and buck-type PFC topologies under BCM control. While the boost converter maintains a more continuous and sinusoidal input current waveform—as depicted in the illustrative results of Figure 8—the buck converter, as shown in the experimental waveforms of Figure 11, exhibits significant distortion near the zero-crossing regions. This contrast underscores the inherent trade-offs associated with the selection of PFC topologies in AC–DC conversion applications.

In addition, based on the measured average input power of 11.72 W and average output power of 10.8 W, the overall drive circuit efficiency is 92.2%. Furthermore, Figure 12 presents a photograph of the experimental setup, in which the prototype driver circuit is powered by an AC source to drive the UV-C LED sterilization lamp. A digital oscilloscope, equipped with voltage and current probes, is employed to observe and record the voltage and current waveforms of the prototype circuit under operation.

Figure 13 illustrates the loss distribution of the prototype UV-C LED sterilization lamp driver circuit, providing experimental validation of the efficiency measurements presented in this study. As shown in the figure, the primary sources of power loss are distributed among several key components. The power switch S₁ contributes 17.6% of the total loss, while diodes D₁–D₄ each account for approximately 2.15%. In addition, diodes D₅ and D₆ exhibit higher loss proportions of 6.86% and 11.45%, respectively. The remaining 55.49% of the total power loss is attributed to other circuit elements and parasitic effects, including magnetic components, capacitors, and the auxiliary circuitry associated with the controller IC. This detailed loss breakdown offers valuable insight into the dominant inefficiency mechanisms within the system and identifies critical targets for future performance optimization.

Table 5. Comparisons between the existing single-stage single-switch AC-DC LED drivers in [31,32] and the proposed one.

Item	Existing Single-Stage Single-Switch AC-DC LED Driver in Reference [31]	Existing Single-Stage Single-Switch AC-DC LED Driver in Reference [32]	Proposed Single-Stage Single-Switch AC-DC LED Driver
Circuit Topology	Combining an isolated SEPIC converter with a lossless snubber	Combining a Vienna converter with an isolated SEPIC-flyback converter	Combining a buck PFC converter with a flyback converter
Input AC Voltage	90~240 V	220 V	110 V
Output Power	18 W (30 V/0.6 A)	18 W (30 V/0.6 A)	10.8 W (90 V/0.12 A)
Number of Required Power Switches	1	1	1
Number of Required Capacitors	4	4	3
Number of Required Magnetic Components	4	3	2
Number of Required Diodes	7	3	6
Measured Power Factor	0.97 at 110 V	0.97	0.9164
Measured Circuit Efficiency	93% at 110 V	95%	92.2%

presents a comprehensive comparative analysis of three single-stage, single-switch AC–DC LED driver circuits, comprising the driver proposed in this study and two previously reported designs [31,32]. The driver described in [31] integrates an isolated SEPIC converter with a lossless snubber and is designed to deliver an 18 W output (60 V/0.3 A) from an input AC voltage range of 90–240 V. In contrast, the topology in [32] combines a Vienna converter with an isolated SEPIC–flyback converter to provide an 18 W output (30 V/0.6 A) from a 220 V AC input. The proposed design adopts a hybrid configuration that integrates a buck-type PFC converter with a flyback converter, yielding a rated output of 18 W (90 V/0.12 A) under a 110 V AC input condition.

The comparison encompasses key performance and design parameters, including circuit topology, input voltage range, output power, number of active switches, capacitors, magnetic components, diodes, measured power factor, and overall efficiency. As shown in Table 5, all three drivers employ a single active power switch and deliver comparable output power levels (approximately 10–18 W) suitable for LED lighting applications. Relative to the circuit reported in [31], the proposed driver achieves a comparable conversion efficiency while reducing the total component count—specifically, one fewer capacitor, two fewer magnetic components, and one fewer diode—thereby enabling a more compact and cost-effective implementation. Compared with the design in [32], the proposed circuit requires a greater number of diodes and exhibits slightly lower efficiency; however, it still achieves a reduction in passive components, using one fewer capacitor and one fewer magnetic element.

Although the designs in [31,32] inherently benefit from the PFC characteristics of SEPIC-based and SEPIC–flyback topologies, respectively, the proposed buck-type PFC driver exhibits a comparatively lower power factor. Nonetheless, this design demonstrates significant advantages in terms of component reduction, achieving an effective balance between circuit simplicity and energy conversion performance.

5. Conclusions

This paper has presented a high-efficiency UV-C LED sterilization lamp driver circuit for AC voltage supply, employing BCM control with integrated power factor correction. Compared with traditional mercury-based UV lamps, UV-C LEDs emit short-wavelength radiation capable of disrupting the DNA or RNA structure of microbial cells, thereby inhibiting replication and achieving effective sterilization. The proposed single-stage, single-switch buck–flyback topology reduces component count, recovers transformer leakage energy, and enhances overall energy conversion efficiency compared to conventional two-stage designs. A 10.8 W prototype was implemented and experimentally evaluated. Under a 110 V RMS AC input at rated output, the driver circuit demonstrated a power factor exceeding 0.9, an efficiency greater than 92%, an output voltage ripple below 2%, and an output current ripple below 5%, confirming the effectiveness, compactness, and cost-efficiency of the proposed design for UV-C LED sterilization applications.

In summary, this paper presents a single-switch UV-C LED sterilization lamp driver circuit that integrates a buck PFC converter and a flyback converter into a single-stage, single-switch buck–flyback AC–DC power converter with input-current shaping capability, achieved through the implementation of a BCM control scheme. Furthermore, the proposed driver incorporates an energy recovery mechanism for the transformer’s leakage inductance, thereby enhancing overall conversion efficiency. The circuit operation is analyzed based on three distinct operating modes under the BCM control strategy. This design builds upon the foundational principles established in [28,29], which employ a DCM control scheme with four operational modes while utilizing the same main circuit topology to achieve PFC. In contrast to the methods presented in [28,29], which differ in both control strategy and circuit parameter design, the proposed approach achieves a threefold increase in output power (10.8 W versus 3.6 W) by delivering a threefold higher output current (0.12 A versus 0.04 A) at the same output voltage of 90 V. Moreover, the proposed driver exhibits a marginal improvement in conversion efficiency (92.2% versus 91.85%), demonstrating its superior overall performance.