While most domestic appliances and office equipment items are plugged directly into wall outlets and powered from high-voltage alternating current (AC), nearly all of their internal circuitry requires a low-voltage direct current (DC). Accordingly, power supplies are required to convert AC voltage to low DC voltage. According to the research by Ecos Consulting, roughly 3 billion AC/DC power supplies are currently used in the U.S. and about 10 billion are used globally [1, 2].

As power supplies become more pervasive, their impact on the environment in terms of power supply energy efficiency has attracted growing international attention. As a first step, a number of efforts have been made to improve efficiency when the appliance is in standby mode [3,6].

Enlarge this picture Fig. 2. Operation waveforms of the proposed method.

However, many of the technical approaches used to reduce the standby power consumption do not necessarily improve active-mode efficiency. Research by Ecos Consulting shows that about 73 percent of the total energy passing through the power supply occurs when the products are in active mode. This implies that there are more opportunities to save energy in the active mode [1, 2].

Thus, governmental and standardization organizations around the world, committed to protecting natural resources, have endorsed measures to encourage the adoption of energy efficient power supplies. The California Energy Commission (CEC) has proposed mandatory efficiency standards for external power supplies. Other regions of the world, which currently depend on voluntary regulation programs, are also considering mandatory standards to encourage higher efficiency of power supplies [7].

Enlarge this picture Table 1. Energy efficiency regulation for active mode

One thing that should be noted in the new CEC standards is that they regulate efficiency at full load and also at partial load. They regulate the average of efficiencies at 25 percent, 50 percent, 75 percent and 100 percent of rated output power, as shown in Table 1.

Typically, the efficiency drops as the output load decreases, because the switching loss becomes dominant at a light-load condition, which sometimes becomes an obstacle to meet the active-mode efficiency regulation. The simplest way to reduce the switching loss is to employ a quasi-resonant operation, which is accomplished by operating the converter in the critical conduction mode (CCM), that is, at the boundary of the CCM and discontinuous conduction modes (CCM/DCM). However, the critical conduction mode operation causes a wide range of frequency variation as the load changes. Another drawback is the increased conduction loss compared to the CCM operation, since the critical conduction mode operation requires a relatively small inductor. The conduction loss becomes more severe when applied to a universal input voltage application, since a smaller inductor is required to guarantee the critical conduction mode operation over the entire input-voltage range.

The proposed hybrid control scheme allows the converter to operate in the conventional hard-switched pulse-width-modulated (PWM) mode when the converter operates in CCM, and operates in the quasi-resonant mode when operating in DCM. Thus, the proposed control scheme can enable quasi-resonant operation only when the converter operates at a medium- or light-load condition. This permits a relatively large inductor from the conventional PWM converter design and improves efficiency by reducing the switching loss at a light- to medium-load condition without increased conduction loss.

Hybrid controller

Enlarge this picture Table 2. Experimental flyback converter specifications. (Vin = 90~265 VAC, Vo = 5 V/6 A).

The quasi-resonant flyback converter topology can be derived from a conventional square-wave, PWM, flyback converter by operating the converter in the critical conduction mode. Because the MOSFET is turned on when the drain-source voltage reaches its minimum value by resonance, quasi-resonant operation reduces electromagnetic interference (EMI) while increasing power-conversion efficiency.

However, the CCM operation undergoes a wide range of frequency variation as the load changes. As the load decreases, the switch’s “on-time” should decrease to reduce the peak drain current and therefore increase the switching frequency. This results in severe switching losses as frequency increases at a light-load condition. Another drawback is the increased conduction loss compared to CCM operation, since the critical conduction mode operation requires a relatively small inductor. The conduction loss becomes more severe when applied to a universal input voltage application, since a smaller inductor is required to guarantee critical conduction mode operation over the entire input voltage range.

To solve the drawbacks of the conventional quasi-resonant operation, the proposed hybrid control scheme combines the conventional PWM operation and quasi-resonant operation. The operation mode is adaptively selected, which allows the converter to operate in the conventional hard-switched PWM mode when the converter operates in CCM, while in quasi-resonant operation when operating in DCM. Fig. 1 shows the simplified schematic of the flyback converter employing the proposed control method. In order to monitor the drain voltage waveforms, the auxiliary winding voltage is used.

Fig. 2 shows operation waveforms of the proposed control method for different load conditions. When the converter operates in CCM, the auxiliary winding voltage is clamped at Vo× Na /Ns until T1, and no oscillation of the auxiliary winding voltage is observed within the minimum switching period (Tsmin). Then, the controller performs the conventional PWM control by turning on the MOSFET at T1 (Operation A). Meanwhile, as the load decreases and the converter enters into DCM, oscillation of the auxiliary winding voltage is observed within the minimum switching period (Tsmin). Then, the controller allows the converter to operate in valley switched DCM (quasi-resonant operation). In DCM operation, the controller finds the first minimum drain voltage within the detection time window (TW) by monitoring the auxiliary winding voltage and turns on the MOSFET with minimum voltage (Operations B and C). Thus, the switching frequency is limited regardless of the operation mode as

Enlarge this picture

Where, TR is the resonance period between the primary side inductance and MOSFET output capacitance.

The above equation illustrates another benefit of the proposed controller. Normally, when the switching frequency decreases, a bigger core is needed in the Switch Mode Power Supply to avoid core saturation. Therefore, it is not easy to select an optimal core size when switching frequency varies widely like a conventional QR converter, and sometimes the selection of a bigger core is inevitable. Because the switching frequency variation is quite narrow due to the proposed scheme as described in Fig. 3, the transformer design becomes quite simple and cost effective, which is the advantage of a fixed-frequency flyback converter.

Test results

Enlarge this picture Fig. 3. Narrow switching frequency variation due to the proposed controller.

In order to demonstrate the validity of the proposed method, three 90~265 VAC/5 V, 30 W off-line flyback converters with different control methods were built and tested according to the specifications shown in Table 2. The same transformer inductance values are used for all cases.

For case A, the switching frequency is fixed and the converter operates in CCM for low-line voltage (Vin = 115 VAC) and in DCM for high line voltage (Vin = 230 VAC). Because the switching frequency is fixed, the peak drain currents are almost the same for low-line and high-line voltages.

Meanwhile, for case B, the converter always operates in critical conduction mode. The switching frequency decrease for low-line voltage, and therefore its peak drain current, is higher than in the other two cases. This increases the conduction loss, causing a decrease of efficiency at low line.

Enlarge this picture Fig. 4. Efficiency comparison when Vin = 115Vac.

For the proposed control method, the converter operates in CCM with maximum switching frequency at low-line voltage while in valley switching DCM for high-line voltage. Thus, the peak drain current for low line voltage is the same as the case A.

Fig. 4 shows the efficiencies of the three prototype converters when input voltages are 115 VAC. As observed, the conventional quasi-resonant operation (case B) shows deteriorated efficiency for a full-load and low-line condition. This is because the reduced switching loss is negligible and increased conduction loss is dominant for full load and low line condition. However, the proposed method shows better efficiency at low-line and full load since it permits quasi-resonant switching only when the converter enters into DCM as the load or input voltage varies. It does not have an increased conduction loss caused by relatively high drain current to guarantee the critical conduction operation over the entire operation range.

Enlarge this picture Fig. 5. Efficiency comparison when Vin = 230Vac.

Fig. 5 shows the efficiencies of the three prototype converters when input voltages are 230 VAC. As can be seen, the proposed method also shows increased efficiency over the other two methods. The conventional quasi-resonant operation shows lower efficiency, since the increased switching frequency for high line increases the switching loss.

Fig. 4 and Fig. 5 shows that the proposed controller is very effective in improving the active-mode efficiency both in high line and low line conditions.

Conclusion

A new hybrid control method can combine a conventional PWM operation and a quasi-resonant operation. The proposed approach reduces the switching loss and improves efficiency at medium- and light-load condition without increasing the conduction loss. The proposed scheme was verified through experimental 90~265 VAC/5 V, 30 W off-line flyback converters.

References

1. Ecos Consulting, “Power Supply Efficiency: What Have We Learned?” Feb. 2004, prepared for the California Energy Commission’s PIER program.
2. Chris Calwell, Arshad Mansoor, Robert Keefe, “Active Mode Power Supply Efficiency: Key Issues, Measured Data and the Design Competition Opportunity,” APEC 2004.
  3. J.P. Ross and Alan Meier, “Whole-House Measurements of Standby Power Consumption,” International Conference on Energy Efficiency in Appliances, September 2000.
  4. Alan Meier, Wolfgang Huber and Karen Rosen, “Reducing Leaking Electricity to 1 Watt,” The 1998 ACEEE Summer Study on Energy Efficiency in Buildings, August 1998.
  5. Yoh Matsushita, “Design for Low Electric Power During Standby State of Fax-Copier Machine,” International Symposium On Environmentally Conscious Design and Inverse Manufacturing, pp. 391 – 395, 1999.
  6. Hangseok Choi, D. Y. Huh “Techniques to minimize Power Consumption of SMPS in Standby Mode,” IEEE Power Electronics Specialist Conference, 2005.
  7. California Energy Commission “Appliance efficiency regulation (CEC-400-2006-002)” January 2006.