Smart grid technology blends communication technologies with the power grid to create a whole new smarter grid that intelligently improves efficiency and reliability.
David Andeen, segment manager, Smart Grid, Maxim integrated, an American company that designs, manufactures and sells analogue and mixed-signal semiconductor products, spoke to Dilin Anand and Ashwin Gopinath of eFY about the smart grid and how to design robust next-gen solutions for it
Q. What are the main areas of focus for the smart grid, and why?
A. The primary focus is on building more functional, integrated smart meters. Smart meters are the volumes and revenue driver in smart grid growth right now, which, in turn, drives the innovation. We envision a single chip providing metrology, security and communication. We are already seeing development of a family of products.
Q. What are the hardware design challenges while integrating such a multitude of functionalities?
A. This question drives to the heart of analogue integration. Integrating multiple analogue functions into one chip is a tremendously diffcult task because it involves merging functions that were originally optimised in different process geometries. Ultimately, the biggest challenge is in making the performance of analogue functions similar or better. In smart meter chips, metrology is an analogue function where we accurately detect energy across a wide current range.
Q. What does a smart grid solution need to become competitive?
A. Smart grid solutions need to provide the highest-accuracy metrology or measurement of the energy. We do this in electricity meters and in distribution automation equipment. We combine the highest accuracy with a high level of analogue and digital integration to produce metrology SoCs.
The next generation of SoCs will feature security and communication. Security is a real need in a smart grid, which is currently not adequately addressed. Metrology and security make these products the most competitive.
Q. What communication technologies will be implemented into metrology SoCs?
A. The current focus is on powerline communication, where we have defined G3 and we are also implementing additional solutions. We have also partnered with RadioPulse for ZigBee solutions. In the future, we will implement additional protocols. We see integration as the primary path for all communication protocols.
Q. For design engineers planning to implement a metrology SoC in their products, what do you suggest to shorten the time to market?
A. Design engineers have a tough job. In addition to hardware, smart meters require a lot of firmware and software for complete operation. These devices have gone from being high-performance meters to being network nodes that still require high-performance metrology.
Design engineers need to convince themselves that they are implementing the best possible solution. They need to read application notes and firmware documents to be sure that they are completely familiar with the system.
It is difficult to give advice on shortening a process that requires a lot of time and effort. The best way to save time is to make good decisions. More time spent making a decision is almost always less than the time required to change a decision.
Q. What are the customer demands?
A. Great question, because smart grid is not just about supplying parts, it is about providing solutions. There are numerous issues in metering, smart metering and distribution automation applications, including magnetic interference/immunity challenges, EMI and others, all the way down to providing solid reference solutions to accelerate the design process.
A strong applications team is a big advantage, which sets a solution provider apart. Furthermore, you need a number of solutions (power supplies, RTCs, isolation products and transformer drivers) that surround the major system SoCs for various meter and smart grid systems. These aren’t just off-the-shelf analogue components, rather they are high-performance analogue products developed for smart grid applications.
Q. Could you give an example of how you were able to solve the emi challenge in a smart grid application?
A. Many of these are proprietary to specific customers. One example, though, is from Germany, where meters must pass a test involving a massive magnet. This magnet is so strong that it immediately erases your credit cards and your computer hard drive. We had trouble even buying the magnet. In these types of cases, our applications engineers find signals buried deep within the noise. They make the solutions work, despite requirements that meters will likely never see in the field.
Q. What functions can a system design engineer expect from next-gen SoC solutions for smart grid?
A. Our next highly integrated product is Zeus, which integrates our highest-performance metrology with hardware security and a 32-bit ARM core for applications and communication stacks. The security in this product is really differentiated because it is not just software security. We are implementing a lot of the features that we put into financia terminals.
Q. What kinds of applications require a high-performance SoC, and what applications would be better off a lower-performance SoC to achieve pricing targets?
A. The ideal application for a high-performance SoC is a high-volume platform. If the platform needs to be leveraged across multiple regions, that’s even better because firmwar changes may be all that is needed. Lower-volume applications probably do better with performance building blocks or limited-function SoCs. If the region you are targeting has not selected a communication protocol, then it is probably safer to pick the building blocks.
Q. How are the current smart grid setups made secure? How can design engineers improve upon the security features when using the new SoCs?
A. There are a variety of approaches in smart meter security today. The most common is tamper protection on the meter, followed by AES-128 encryption on the data being communicated to and from the meters. In this type of configuration there are still gaps, such as the data path from the metrology to the communication hardware. If those lines can be sniffed or modified, then security has been breached.
Our recommendation is to secure the life-cycle of the product. Purchase through a trusted supply channel. Ensure authentication in the manufacturing process. This can be done with integrated or discrete products. Ensure authentication during the installation process as well. Finally, close gaps on the meter itself. For the example above, either select an integrated solution or bury the lines in the PCB and secure them separately, prior to routing to the communication module.
These solutions are much more difficult than this discussion warrants, but security is absolutely critical to our energy infrastructure and we need to build it into these systems, rather than patch it up later. That is definitely the best approach.