|
Synchronous buck converters, as opposed to conventional buck converters, can achieve high efficiency in today's low-voltage, high-current applications because they replace the catch diode of buck converters with a MOSFET. As a result, the power they dissipate in the off-period is reduced significantly.
On the other hand, use of a MOSFET (two required in the traditional configuration) in place of the catch diode
prohibits the converter from entering the discontinuous conduction mode (DCM), and thus degrades the efficiency at light load. Optimizing the converter's associated component circuitry is also difficult. How do we address these drawbacks?
Hardware considerations
With two MOSFETs in the synchronous buck converter, we need to avoid the shoot-through current, which effectively shorts the power supply to the ground through the MOSFETs. This current introduces a large switching loss, and in the worst case the MOSFETs or the power supply can be damaged. To get around this problem, we introduce a suitable delay, known as dead-time, during the switching cycle. On the other hand, for high conversion ratios, i.e., when the difference between input and output voltage is large, the conduction time for the two MOSFETs are different. In some cases, the difference can be more than a factor of 10. For example, a converter with a 24-volt input and 1.8-volt output requires a duty ratio of 7.5 percent, so the ratio of the conduction time is 0.075:0.925, which is about 12 times. The two MOSFETs should therefore be appropriately rated for power capability to optimize the design in terms of component cost.
These aforementioned issues underscore the difficulty of building synchronous buck converters with discrete components. Securing MOSFETs with the ideal turn-on to turn-off time and determining the proper dead-time and the characteristics of the related circuitry consumes considerable engineering effort. In particular, the designer often runs up against a limited choice of MOSFETs when it comes to finding devices with suitable electrical parameters for a given power rating.
Standby efficiency
While using a MOSFET in place of the catch diode reduces the conduction loss, it allows bidirectional flow of the inductor current. Thus, the synchronous buck converter maintains operation in its continuous conduction mode (CCM), versus the discontinuous conduction mode (DCM), even at light load. Figs. 1a and 1b show the inductor current of a buck converter and a synchronous buck converter both for high and low output current. Both converters maintain CCM operation at high output current. At light load, the traditional buck converter, with a diode in the circuit, goes to DCM operation because the diode blocks negative current in the inductor.
Figure 1a: Inductor current in a buck converter
Figure 1b: Inductor current in a synchronous converter
Here we see that CCM operation at light load (implies low output power) in synchronous converters is a drawback if standby efficiency is a major concern in your application. In short, the MOSFET's conduction loss comes into play, and the total power dissipation will be relatively large. In DCM operation, there is no conduction loss when the inductor current is zero. In addition, zero-current switching operation helps reduce switching loss. To summarize, synchronous buck converters yield high efficiency at high output current but low efficiency at low output power. We need to address this issue especially if the converter is in standby mode during much of the time.
We can do this by adding control circuitry that detects the current through the MOSFET and turns it off when the current is zero in order to block any negative inductor current. Hence, the MOSFET effectively functions as a diode that enables the synchronous buck converter to operate in the DCM while maintaining low voltage drop at turn-on. The added circuitry to achieve this objective will drastically improve the efficiency at low output power, and not affect the converter's performance when output current is high.
Single-chip solution
The single-chip solution is the most effective way to implement a good synchronous buck converter. It optimizes the size of MOSFETs to achieve the best switching ratio. The MOSFETs are integrated, and so designers can optimize dead-time control and driving circuitry. Moreover, it's easier to detect MOSFET current and turn off the MOSFET to achieve DCM operation.
Consider, for example, the LM310x family of synchronous buck converters. They have internal MOSFETs and all circuitry to implement a synchronous buck converter. In particular, they include a zero coil-current detection circuit that enables DCM operation. Fig. 2 shows a typical application for the LM3102.
(Click on Image to Enlarge)
Figure 2: High-efficiency synchronous buck converter using the LM3102
Figure 3 shows the efficiency curves for the LM3102 and a buck converter (non-synchronous) with an output voltage of 1.8 volts. The LM3102 achieves an efficiency (for the same input voltage) that's 10 to 15 percent higher than for a buck converter.
Figure 3: Comparing efficiency, LM3102 versus non-synchronous buck converter
Similarly, consider the improvement in efficiency for a 12-volt in, 1-volt out converter using the LM3100, versus a conventional synchronous buck converter (Fig. 4). Both converters have an efficiency of greater than 70 percent for high output current. At light load, however, the efficiency of the conventional synchronous converter drops, apparently due to CCM operation. The converter built around the LM3100, however, can enter the DCM to maintain high efficiency at low output current.
Figure 4: Comparing efficiency, LM3100 versus traditional synchronous buck converter
About the authors
L. K. Wong is a senior product application engineer, and T. K. Man is a product application manager, at National Semiconductor's Hong Kong Power Management Design Center.
|
