Systematic Efficiency Improvement Turns Power Designs Green

Emerging global emphasis on eco-friendly design has brought efficient power electronics closer to the center-stage of system design considerations. Power electronics professionals have to respond by learning and adapting to newer/better ways of doing their jobs to be more effective and more efficient. The best way to achieve this goal is a hierarchical design approach that enables systematic solution. The foundation is laid by strong product definition comprising of clear specifications and an appropriate architecture. The next level is the complete product design using appropriate toolkit (e.g. simulations, prototyping, analyses, etc). The final layer is product validation and testing.

DRIVERS FOR HIGH EFFICIENCY REQUIREMENT
The primary driver for any paradigm change is usually the economics, and that is the case for high efficiency as well. The need for smaller, lighter and quieter electronics can only be fulfilled by high efficiency power conversion. In recent years, these have been complemented nicely by the eco-awareness that has led to regulatory requirements for power savings spanning geographies and applications. As a result, there is a compounding drive effect of what can be called an Eco2 (Economics x Eco-awareness).

With all these external incentives, there is a challenge posed to the power electronics community on how to come up with high efficiency power conversion solutions which are economical, sustainable, scalable and repeatable. In other words, is there a systematic/holistic approach to designing for high efficiency?

Traditionally, a multi-disciplinary field like power electronics has not followed a set development approach. However, as standardization of requirements and form factors takes foothold, there has to be a convergence in development approaches also. Any development involves four distinct phases: specifications/concept development, System level design/architecture, detail design and testing/refinement. By adapting a hierarchical approach, where the initial phases are handled in modular fashion and latter stages in an integral method, the power supply designers can achieve the best success for high-efficiency projects.

CONCEPT DEVELOPMENT
This phase forms the foundation of the design and is often treated lightly – leading to a drawn out development cycle with unsatisfactory results. When it comes to high efficiency designs, it is critical to understand the operating conditions where high efficiency is required. Often, it is only for full load, low input voltage conditions. However, if the requirement is regulatory standards based, it is more likely to cover a broader load range. Generally, the regulatory requirements are specified at nominal input voltage only.

Often, the relationship between the specifications and efficiency is not obvious, but requires more exploration in order to guarantee best efficiency performance. Once specifications are clear and agreed on, the next task is to design the system for the best efficiency at a reasonable cost.

SYSTEMATIC IMPROVEMENT OF POWER SUPPLY EFFICIENCY
Three methods can be employed to increase the power supply efficiency:
• System architecture considerations
• Topology improvements
• Component optimizations

System architecture
System architecture is the critical phase in design and is often ignored. Many companies use their design engineers as “de facto” system architects, but without giving them the required tools and mandate to perform this role. The architecture decision is not limited to choosing the power topology alone – it also involves decisions such as power stage partitioning, voltage levels for each stage, understanding and ensuring regulatory compliance etc. Traditionally, going to a new architecture involves a certain amount of risk and only a handful of companies have taken that risk with varying level of rewards. While some new architectures focus on efficiency improvements without cost considerations, others focus solely on cost reductions. Ideally, a combination of both cost and performance improvements is required for success.

Topology selection
Choosing the right topology for a given application is a very important decision. Traditional topologies such as buck, boost, buck-boost (for non-isolated applications) and flyback, forward, push-pull, half-bridge (for isolated applications) have served the power designers well over the years, but in the context of efficiency requirements, they may fall short on performance. In these circumstances, it is worth considering going to some emerging soft-switching topologies for a given application. However, blindly going to a new topology without fully considering its suitability to an application will likely backfire. It is useful to refer to the topology-application grid (Figure 1) where different applications and power level are shown to correspond to selected traditional and emerging topologies. Certain topologies are better for wide input range, while some others are better for restricted load range. Some topologies are more conducive to synchronous rectification, others allow better cross-regulation for multiple outputs. Thus, a careful evaluation of the options for a given application is necessary.


Figure 1: Application topology grid - a structured topology selection tool.
Click to enlarge

Component optimization
Once system architecture and topology are selected, the design involves selecting the right components (including the magnetic elements). As in any multi-variable problem, some variables have to be fixed before proceeding to select the other variables. In power supply design, it is very helpful to identify a critical parameter that can have maximum impact on the results/performance for a required metric (e.g. efficiency). This design approach is called pivot point approach. For example, in a flyback converter design, the primary to secondary turns ratio is the pivot point and if that is selected without due consideration, it is very difficult to optimize the design. A higher turns ratio reduces the primary circulating current and secondary diode stress. Both these are important for improved efficiency. On the other hand, the higher turns ratio also increases the leakage inductance and losses in the clamping circuits. Thus, choice of the turns ratio is also related to the switching frequency choice (at lower frequencies, the turns ratio can be pushed a little higher while at high frequencies, the trade-off requires lower turns ratio). Similar to flyback converter, pivot points can be identified and used as starting points for designs using other topologies also.

While each of these methods can have an impact, a judicious and prioritized combination of the three can yield the best results. In terms of effort required, component optimization may be the path of least resistance, but without a good architecture and with choice of an inappropriate topology, it can lead to meagre returns for the cost incurred.

Once the pivot point has been designed, there is additional optimization possible with certain power stage component designs. For example, a cost vs. performance trade-off can be made by selecting current-sense resistors (low cost, higher losses) or current sense transformers (higher cost, low losses). In some cases, the PWM controller selection drives choices of many other components and is critical for better efficiency. Many PWM controllers with integrated high-voltage start-up circuit eliminate high power consumption in bias start-up resistors and thus reduce standby losses.

With the advent in semiconductors, there are many novel choices available that further help improve the efficiency while reducing the cost. In continuous conduction mode (CCM) PFC circuits, the boost diode is widely recognized as the major contributor of losses – directly and indirectly. Recent technology improvements have allowed great improvement in this realm.


Figure 2: Efficiency improvement through judicious component choices.
Click to enlarge

The need for structured design methodology in power supply design is highlighted in this article. This structured approach involves multiple phases including specifications, architecture, design and validation. When this hierarchical approach is applied to power supply designs, it becomes easier to achieve high efficiency in a time and cost effective manner.