4th Power Analysis & Design Symposium 2015

Early DSP-based power converters were mostly used to “imitate” proven analog compensation filters in combination with programmability and design flexibility. These devices were able to support different converter types and topologies and enabled designers to quickly respond to custom specific requirements with no or very little re-engineering and qualification efforts.

However, Digital Control is much more than just software-based analog circuit emulation. With increasing acceptance and experience in various applications and markets, digital control has entered a new level of non-linear control. Over the years adaptive and predictive control methods have been developed to increase efficiency, performance and reliability simultaneously.

This lecture introduces one key example of how the flexibility of digital control loops can help to improve the total performance and bypass/solve major common design tradeoffs in peak current mode controlled power converters using Adaptive Software Slope Compensation (ASSC). The basic concept, implementation and test results will be discussed and verified using OMICRON Lab’s Bode 100.

The loop gain measurement is a standard method to verify the control loop stability of voltage regulators and switching power supplies. However, like for all measurements certain requirements and parameters need to be met to ensure correct and trustworthy results. The choice of the injection point for example is critical and defines the limits of the voltage loop gain measurement. Furthermore the signal level of the injected signal is crucial.

This lecture focuses on the loop gain measurement method and its application using the Vector Network Analyzer Bode 100.  In addition to the voltage loop gain measurement two additional measurement methods are outlined that can provide results even when no suitable injection point for the voltage loop gain measurement is available.

According to the latest market research, the digital power phenomenon is about to hit us like thunder.

Most larger corporations in the power industry have already embraced this new technology, and are introducing products with performances unimaginable only a few years ago. New efficiency regulations are also increasingly forcing companies to swallow the digital power pill as such high performances are seldom possible with simple analog solutions. Whilst many medium to small size companies are aware of the changing landscape of the power market, they do not have the man power nor the resources to invest in training their engineers in this new technology.  

This session is aimed at addressing the above issue. The most challenging part of a digital power design is the discrete time control loop coefficient calculations. In this technical session we will discuss this process in detail and highlight the calculations that you need to make in order to design a stable digital control loop. Having covered the foundation of digital power control loop design in detail, we will then introduce arguably the world's most sophisticated digital power supply design software which automatically calculates the control loop coefficients for the Infineon, Microchip and Texas Instruments range of microprocessors. The session concludes with live demonstration of the digital power control loop that we designed during the session, its real life loop measurement and direct comparison with the theory. 

The gap between calculation and the real world can be huge. This especially applies for the world of electronics. Deviations between theoretical filter design and the achieved performance of the filter can be caused by many reasons. For example unknown parasitics of the used discrete components or direct crosstalk in the printed circuit board layout can strongly influence the over-all filter characteristic.

So, there are a view things to know, to get the filter working as it should. This presentation outlines common pitfalls and helps to remember all these effects and to avoid mistakes in future designs.

Debugging electromagnetic interference (EMI) is a challenging task. Identifying and mitigating the source(s) of unwanted emissions is often iterative and time consuming, multiple test rounds at an EMC laboratory costly.

The high sensitivity and powerful spectrum analysis capabilities of the Rohde & Schwarz digital oscilloscopes RTE and RTO allow to do a lot of the analysis already in the lab. Using near-field probes and coupling networks an R&D engineer can identify hot spots and do pre-qualification before going to the EMC test house. EMI reduction measures can easily be tested. This session gives an overview on the EMI debugging capabilities of the Rohde & Schwarz RTE and RTO oscilloscopes as well as an hands-on guide how to approach the problem.

In today´s world, R&D engineers face shorter and shorter design cycles, high cost/innovation pressure and increased design complexity. Computer-aided design is essential to keep pace with constantly growing requirements. LTspice is the world´s most popular power electronics simulator. It aims to simplify power supply design and verification tasks, leaving more time for creativity and innovation. 

This session gives a comprehensive overview about the various aspects of SMPS design (topology selection, impact of parasitics/layout, bode/impedance analysis, thermal simulation, etc) that can be covered with LTspice. 

Many engineers are familiar with the 1-port reflection impedance measurement and the 2-port shunt thru measurement.  The former is ideal for measuring impedance in the range of 1 ? to a few k?’s while the latter is ideal for measuring from 1 m? to a few ?’s.  

The 3-port measurement can be used to measure from a few m?’s to a few k?’s and is also ideal for measuring the input impedance of switching power supplies and other system level circuits. 

This presentation provides an introduction as well as detailed instructions for the setup, calibration and examples using the 3-port impedance measurement.