1. Introduction
Digital controllers are currently popular for high-frequency, low-power switching mode power supplies because of their advantages in terms of programmability, different implementation and low sensitivity to variations [1]-[4]. An LED driver controlled by a digital chip allows the control strategy to be user-defined, and therefore the circuit design becomes more flexible and better able to fulfill our target requirement. In the proposed circuit, we program a low-cost microcontroller to achieve closed loop control and Pulse-Width-Modulation (PWM) functions. Compared with the complicated analog circuit that utilizes a comparator and optical coupler to close the voltage and current-control loop, the proposed digital circuit only requires the user to define the voltage and current control by code, which is not only simpler but also reduces the material cost. LED drive circuits often contain power factor correction (PFC) functions. A single-stage topology such as Flyback or boost which is switched by the PWM signal can have PFC function with good power factor. PWM signal could control the switch to make the input current in phase with the input voltage. LED driving using a single-stage PFC circuit often has a huge light flicker, which can be harmful to eyes [5-7]. This flicker is triggered by the twice-line-frequency driving current. According to [8] and [9], another switch added to the secondary side to control the power flow will reduce the capacitor’s value. However, this is not suitable for a condition that allows only one switch. Our circuit implements a pre-stage, which helps ensure that the driving-current ripple is low, even with a small-output capacitor. The benefit here, in comparison to the conventional two-stage PFC, is that it only uses one MOSFET. Meanwhile, the output current ripple is much smaller than that of the conventional Flyback with the same output capacitor. Therefore the proposed circuit will lead to a smaller light flicker.
A new LED driver circuit, as shown in Fig. 1, is proposed in this paper. The output current has a low ripple because of the energy-storage capacitor on the primary side. The proposed circuit is compatible with a digital controller. The operating principle of the proposed circuit was analyzed, a theoretical analysis of the output-current ripple was conducted, a design procedure for the proposed LED driver circuit was provided, and a prototype was built to verify performance.
Fig. 1.Circuit topology of the proposed single-stage
2. Circuit Configuration and Analysis
Fig. 2 demonstrates the operating principles of the proposed circuit. From t0 to t1, the switch Q1 is on, the coupled inductor is charged, and the primary-side capacitor is discharged. The secondary side diode is biased, as is the current flow from the output capacitor to the LED load. From t1 to t2, on the primary side, the coupled inductor is discharged, and the primary-side capacitor is charged. Current flows through the secondary-side capacitor and charges the output capacitor. From t2 to t3, no current flows through the primary side. The situation on the secondary side is the same as that of t1 to t2. In the last period, the secondary-side diode is reverse biased. The load current comes only from the output capacitor.
Fig. 2.Operating principles of the proposed circuit
3. Digital Control of the Proposed Driver Circuit
Digital control of the proposed circuit is implemented using the dsPIC30F2020 chip [10]. The system diagram is shown in Fig.3.
Fig. 3.System diagram of the proposed circuit
The circuit works with a dual-loop control, with both the voltage regulation and current regulation. Here the dual loop control is different with the common series connected one. The traditional one is taking the voltage error as the reference of the current loop. We use a parallel structure, as shown in Fig.3. The advantage is that implementation is simple. The voltage and current loop will not have interference with each other. The algorithm of control is presented in Fig. 4.
Fig. 4.PI control algorithm flow chart
The PI controller for the voltage loop can be represented as
in the continuous domain. Through the bilinear z-transformer [11-17], it can be represented as
Since T=10-5s, we can get
Therefore, the control algorithm of voltage loop can be written as
The current-control loop can also be represented in this way. A similar equation can be derived, such as
Then the outputs of the voltage loop and the current loop are compared. The one with the larger value has a larger error, and it should therefore be chosen as the input for the PWM.
4. Analysis of the Current Ripple and Flicker
Methods for analyzing the twice-line-frequency ripple for a PFC circuit are introduced in [18-20]. In this paper, we found and solved the relationship between the twice-line-frequency ripple and the circuit parameters [18]. We first analyzed the output-voltage ripple. We know the current ripple of the LED is related to the voltage ripple of the LED; therefore, as long as we know the voltage ripple, we can analyze the current ripple based on the I-V curve of the LED. The averaged state-space model over Ts can be built based on the voltage-time balance of the coupled inductor and the Flyback transformer. Eqs. (6) and (7) are derived from the averaged state-space model:
Furthermore, we decomposed each state and input into DC and ripple components. Then, we gathered the first order ripple terms in the differential equations. We did not consider the perturbation of the duty ratio because the circuit was designed to operate with a constant duty ratio and a constant switching frequency. Finally, we get Eq. (8), which describes the perturbations in the primary-side capacitor’s voltage and in the output voltage:
The twice-line-frequency ripple components of the output voltage can be solved according to the following steps [18]. First, the equations above can be represented as
Therefore, the ripple dynamic equations of a certain ripple can be written with a superposition of sinusoidal functions on the right-hand side:
where (ω∈{ωd ,2nωL;n =1,2,...},i=1,...,M)
The solution to Eq. (10) is assumed to have the form
and can be solved as
where A is
The inverse of the matrix can be found using MATLAB. We neglect the trivial terms of the parameters and retain the dominant ones. Finally, we get the approximation equation for
Because xS2 is much smaller than xC2 , the twice-line-frequency ripple’s waveform can be approximated as
Where ω is 2π times the twice-line frequency. According to the energy balance of the Flyback transformer, we can derive the equation
Based on this equation, Eq. (14) can be re-written as
According to the I-V curve for the LED, as shown in Fig.5, we can derive the relationship between the current ripple and the voltage ripple as
Fig. 5.LED V-I Characteristics
Where rc is the current ripple rate and rv is the voltage ripple rate.
Once the current ripple has been estimated from the parameters of the circuit, the intensity of the light flicker can also be estimated. The following is the procedure for deriving the relationship between the current ripple and the two measurements of the light flicker, which are the flicker percentage and the flicker index. Fig. 6 shows the luminous flux waveform of the LED. The flicker index is defined as
Fig. 6.Waveform of the luminous flux from the LED
Because the output luminous flux of the LED has an almost sinusoidal shape,
So, Eq. (19) can be expressed as
Because
we get the relationship
5. Design Procedures of the Proposed Circuit
Fig. 7 demonstrates the design procedures for the proposed circuit. In this paper, we mainly discuss the guidelines for choosing the coupled-inductor value, the transformer’s magnetizing inductance value, and the primary-side capacitor’s value. We also analyze the voltage and the current stress for switching devices to acquire the knowledge needed for device selection. First, we calculated the Flyback magnetizing inductance. In order to achieve a good power factor, we needed to guarantee that the coupled inductor operates in a discontinuous conduction mode (DCM), as shown in Fig. 8.
Fig. 7.Design procedures for this proposed circuit
Fig. 8.Current waveform for the coupled inductor
Therefore,
On the other hand, we can get
where Rload is the equivalent resistance of the LED, V1(t) is the primary-side capacitor’s voltage, and Ts is the period of the switching cycle.
Substituting Eqs. (24) and (25) into Eq. (26), we get the Flyback magnetizing inductance:
where Po is the output power. To simplify the calculation, we approximate V1=V1(t). Therefore, to guarantee that LB operates in the DCM for a universal input range, we demand that the constraint
be satisfied. Finally, we get
where Vin_peak@88v is the maximum value of Vin when its Root-Mean-Square (RMS) value is 88 V. In our prototype design, we found that L ≤ 0.77mH . Here, we choose 0.3 mH. Next, we calculated the coupled inductance. The maximum voltage of the primary capacitor should be set to a certain value. Here, we chose a maximum voltage of 450 V. Through derivation, the primary side capacitor’s voltage, V1, can be expressed as
where Vin_peak is the maximum value of Vin. Therefore, we got L/LB=0.75. Third, we calculated the maximum duty ratio. For a Flyback transformer, the maximum duty ratio cannot exceed 0.5. For a universal input range, the maximum duty ratio occurs at 88 V. Because
we found that D=0.25. Fourth, we calculated the turns ratio of the Flyback transformer. To guarantee that L operates in the DCM, we can use
Therefore, we calculated the turns ratio as n = 3 . Fifth, we decided on the capacitance of the primary side capacitor. We fixed the output capacitor at a certain capacitance. In our prototype, considering the package size, we chose the value to be 1880 uF, and we aimed to achieve a current ripple of less than 15%. According to Eq. (18), we can estimate that a current ripple under 15% means a voltage ripple under 3.93%. On the other hand, with averaged state-space modeling, we derive
Where ω is two times the twice-line frequency. Based on Eq. (33), we established that the primary-side capacitor should have a capacitance of at least 10 uF. As to the existence of a high switching-frequency ripple, the value of C1 should be chosen to be larger than the same critical value. Sixth, we found the voltage and current stress of the switching device. Within the analysis, we obtain
where Vsw_max, Vd1_max, Vd2_max, and Vd3_max are the maximum voltages of the switch Q1 and of the diodes D1, D2, and D3; isw_max, id1_max, id2_max, and id3_max are the maximum currents of the switch Q1 and of the diodes D1, D2, and D3; and V1@264V shows the voltage of V1 when the input voltage is 264 V.
6. Simulation Verification
Our simulation verified that the current-ripple rate in the simulation matched our value calculated on the derived equations very well. The simulation results are shown in Fig. 9. Our simulation was designed for various input AC voltages. The simulation parameters were as follows: output voltage Vo=33 V, output power Po=33 w, C1=10 uF, C2=1880 uF, L=0.3 mH, and L/LB=0.75.
Fig. 9.Simulation result of output current ripple versus input AC voltage
7. Experimental Verification
Fig. 10 shows a picture of the hardware prototype we built based on the design guidelines. Fig. 11 shows details of the circuit board. Fig. 12 and Fig. 13 present the experimental results of the conventional Flyback LED driver and proposed LED driver with 110V input voltage.
Fig. 10.Hardware prototype of the proposed circuit
Fig. 11.Details of the proposed circuit board
Fig. 12.Experimental results of the conventional Flyback with 110V input voltage
Fig. 13.Experimental results of the proposed circuit with 110V input voltage
Fig. 14 and Fig. 15 present the experimental results of the conventional Flyback LED driver and proposed LED driver with 220V input voltage . From the results, we can conclude that the proposed LED driver circuit has a much lower current ripple than that of the conventional circuit. Meanwhile, the power factor of the proposed circuit is above 0.90, which is considered to be good.
Fig. 14.Experimental results of the conventional Flyback with 220V input voltage
Fig. 15.Experimental results of the proposed circuit with 220V input voltage
8. Conclusion
We provide a design guideline for the digital-controlled LED driver. An experimental prototype was built based on the design guideline. The experimental results verified that the proposed circuit had a lower current-ripple rate than the conventional Flyback circuit and that it had a power factor above 0.90.
References
- Guo, S, Gao, Y., Xu, Y., Lin-Shi, X., and Allard, B., “Digital PWM controller for high-frequency low-power DC-DC switching mode power supply”, Power Electronics and Motion Control Conference, 2009. IPEMC '09. IEEE 6th International, pp. 1340-1346, 17-20 May 2009.
- Liu, Y. and Sen, P.C., “Digital control of switching power converters”, Proceedings of the 2005 IEEE Conference on Control Applications, vol. 2, Toronto, Canada, August 2005, pp. 635-640.
- Maksimovic, D., Zane, R., and Erickson, R., “Impact of digital control in power electronics”, IEEE Power Semiconductor Devices and ICs, Proceedings 2004, 16th International Symposium, pp. 13-22.
- Ma, L., Wang, X., Sun, D., and Lv, X., “Research and simulation on voltage-feedback type digital stabilized voltage switching mode power supply”, Information Engineering, 2009. ICIE '09. WASE International Conference on, vol. 2, pp. 188-191, 10-11 July 2009.
- Wang, S., Ruan, X., Yao, K., Tan, S-C., Yang, Y., and Ye, X., “A flicker-free electrolytic capacitor-less AC-DC LED driver”, IEEE Transactions on Power Electronics, vol. 27, no. 11, pp. 4540-4548, Nov. 2012. https://doi.org/10.1109/TPEL.2011.2180026
- Wilkins, A., Veitch, J., and Lehman, B., “LED lighting flicker and potential health concerns: IEEE standard PAR1789 update”, Energy Conversion Congress and Exposition (ECCE), 2010 IEEE, pp. 171-178, 12-16 Sept. 2010.
- Lehman, B., Wilkins, A., Berman, S., Poplawski, M., and Miller, N.J., “Proposing measures of flicker in the low frequencies for lighting applications”, Energy Conversion Congress and Exposition (ECCE), 2011 IEEE, pp. 2865-2872, 17-22 Sept. 2011.
- Nagao, M., “A novel one-stage forward-type power-factor-correction circuit”, Power Electronics, IEEE Transactions on, vol. 15, no. 1, pp.103-110, Jan 2000. https://doi.org/10.1109/63.817368
- Zhang, J., Lu, D.D.-C., and Sun, T., “Flyback-based single-stage power-factor-correction scheme with time-multiplexing control”, IEEE Transactions on Industrial Electronics, vol. 57, no. 3, pp. 1041-1049, March 2010. https://doi.org/10.1109/TIE.2009.2028336
- Microchip, “28/44-Pin dsPIC30F1010/202X enhanced flash SMPS 16-bit digital signal controller”, DS 70000178D, 13 Mar. 2014.
- Liu, Y.-F., Meyer, E., and Liu, X., “Recent developments in digital control strategies for DC/DC switching power converters”, IEEE Transactions on Power Electronics, vol. 24, no. 11, pp. 2567-2577, Nov. 2009. https://doi.org/10.1109/TPEL.2009.2030809
- Xie, M., 2003. Digital control for power factor correction. M.S. Thesis, Center for Power Electronics Systems, Virginia Polytechnic Institute and State University, Blacksburg, Virginia, pp. 1-109.
- Suul, J.A., Molinas, M., Norum, L., and Undeland, T., “Tuning of control loops for grid connected voltage source converters”, Power and Energy Conference, 2008. PECon 2008. IEEE 2nd International, pp. 797-802, 1-3 Dec. 2008.
- Olayiwola, A., Sock, B., Zolghadri, M.R., Homaifar, A., Walters, M., and Doss, C., “Digital controller for a boost PFC converter in continuous conduction mode,” in Proceedings 1st IEEE Conf. Ind. Electron. Appl., 2006, p p. 1-8.
- Barai, M., Sengupta, S., and Biswas, J., “Dual-mode multiple-band digital controller for high-frequency DC-DC converter”, IEEE Transactions of Power Electronics, vol. 24, no. 3, pp. 752-766, Mar. 2009. https://doi.org/10.1109/TPEL.2008.2008391
- Corradini, L., Costabeber, A., Mattavelli, P., and Saggini, S., “Parameter independent time-optimal digital control for point-of-load converters”, IEEE Transactions on Power Electronics, vol. 24, no. 10, pp. 2235-2248, Oct. 2009. https://doi.org/10.1109/TPEL.2009.2022397
- Krismer F. and Kolar, J.W., “Accurate small-signal model for the digital control of an automotive bidirectional dual active bridge”, IEEE Transactions on Power Electronics, vol. 24, no. 12, pp. 2756-2768, Dec. 2009. https://doi.org/10.1109/TPEL.2009.2027904
- Choi, J.Y. and Cho, B.H., “Small-signal modeling of single-phase power-factor correcting AC/DC converters: a unified approach”, Power Electronics Specialists Conference, 1998. PESC 98 Record. 29th Annual IEEE, vol. 2, pp. 1351-1357, May 1998.
- Nagao, M., Nakakohara, T., Jinno, M., and Harada, K., “Analysis of high power factor AC-DC boost converter”, T.IEE Japan, vol. 114-D, pp. 1139-1148, 1994.
- Erickson, R., Madigan, M., and Singer, S., "Design of a simple high-power-factor rectifier based on the Flyback converter", Applied Power Electronics Conference and Exposition, 1990. APEC '90, Conference Proceedings 1990, Fifth Annual, pp. 792-801, 11-16 March 1990.