The main market requirements for power supplies, in addition to the basic technical performance specifications, are high efficiency, compact size and reliability. This article demonstrates how today's semiconductor components are being specifically developed to reduce conduction and switching losses, as well as provide noise immunity, that can increase efficiency and reliability even as applications become smaller.
240W power supply application
Figure 1 shows the block diagram of a 240W power supply that has a power-factor corrected front end. This structure is typical for applications ranging from 200W to 600W. This power factor correction, or PFC, stage is driven by an analog controller such as a PFC/PWM combo controller. For this power level, the PFC controller operates in continuous conduction mode. As the PFC controller switches fast, typically at 100 kHz, the switching losses must be low. Additionally, the RDS(ON) or Vce(sat) also must be low to ensure that the losses from switching the conducting boost diode are kept to a minimum. If a MOSFET is used, its body diode does not conduct during normal operation, thus eliminating the need for a fast-body diode. If an SMPS IGBT is used to implement the switch, it is recommended to add an anti-parallel diode to prevent the IGBT from reverse bias, which can occur under transient conditions.
Figure 1: 240W output PFC power supply
The output of the PFC circuit is followed by the main power conversion stage. This power stage is fed by a reasonably well regulated voltage, allowing the designer to optimize the design for a narrow input operating range. This increases the efficiency of the second stage and generally compensates for the loss in efficiency introduced by the PFC stage.
The output topologies typically used in power supplies are either half-bridge or two-switch forward. In the two-switch forward topology, both switches are turned on at the same time. In the half-bridge, they are switched on in anti-phase. In both cases the high-side device needs a special driver. Pulse transformers can be used for this function, but in modern designs, it is normal to use a half-bridge driver IC as these provide simpler, more reliable and more cost effective drive solutions. In the case of the-half-bridge topology, an additional inverter is needed to provide the anti-phase drive. Further, analog or digital dead-time control is needed to prevent simultaneous switching of the devices.
The switches used, invariably MOSFETs, operate differently in each topology. For the two-switch forward solution, the body diodes do not normally conduct. In many kinds of half-bridge solutions, they will conduct, and will be forced to turn off by the other MOSFET. As a result, it is beneficial to use fast-recovery MOSFETs in these applications.
Using higher switching frequencies in power supply applications reduces the size of the PFC choke and of the transformer; however this will also increase the switching losses. This size requirement conflicts with the need for increased efficiency: for a given technology, devices with lower switching losses tend to have worse conduction losses. To achieve low switching losses in MOSFETs the die area needs to be reduced to lower the gate charge, a main determinant of switching losses. This reduced area leads, in turn, to an increased RDS(ON) , which is the criterion determining conduction losses. Increased efficiency in the power semiconductors in a circuit also leads to a more compact solution, as the size of the heatsinks can be reduced. Finally, the use of new MOSFET technologies allows a given RDS(ON) to be put in a smaller package size, further reducing the system size.