Mixed-signal ASICs are prevalent throughout the industrial controller market even though at first glance this market does not exhibit the features people associate with undertaking an ASIC design. This paper discusses how mixed signal ASICs are changing the face of industrial control systems.
Functional safety is a concept applicable to programmable electronic systems across all industrial sectors, and is fundamental to the enabling of complex technology used for safety-related systems and, increasingly, non-safety related systems. For a number of years Swindon has designed systems to meet the requirements of the international safety standard IEC 61508 to ensure the necessary risk reduction required to achieve safety for the equipment.
ASICs are often thought to be only applicable to high production volume applications but, by combining expertise in design with modern foundry manufacturing technology, they have become a cost effective alternative to traditional electronic circuits in a wide spectrum of products ranging from consumer electronics to medical, sensor interfaces and instrumentation. One of the fastest growing application areas is that of industrial control systems.
Whilst industrial controllers cover many industries and come in many forms, they all carry out a similar function: they monitor a variable, compare the measured value with a desired value and adjust a controlled variable to reduce any discrepancy. This usually incorporates some sort of feedback loop with three term (proportional, integral and derivative) control. The signals generated by the sensors that carry out the monitoring function may be digital, but more often are analogue values of temperature, pressure, level or analytical quality, and these need to be conditioned and digitised to allow them to interface with the control systems. The nature of the sensors has changed as material science has allowed the development of CMOS sensing elements, which lend themselves to integration either on chip or in the same package. Industrial sensors, which are often located in confined and difficult environments, are typically in operation for many years and mixed signal (analogue and digital) ASICs offer a rugged and compact solution with the additional benefit of obsolescence protection, an important factor for lifetime maintenance and spares. Communication between the sensor and the controller can be one of many standard or proprietary protocols, either RF or wired. Data rates tend to be low, so the most common protocols are the Serial Peripheral interface, (SPI) or the Inter-Integrated Circuit (I2C).
One of the first areas of industrial control to adopt mixed signal ASICs was motor controllers, allowing low cost variable speed drive systems to be developed. They include functionality for current sensing, control of the motor, encoders to sense the position and speed of the motor, and communications interfaces that enable the motor control logic to link to other parts of system. Variable speed drive systems optimise the speed of a motor to the duty required and are widely used to control pumps to give a constant head or constant flow. They have replaced flow control valves in many bulk flow applications like water distribution giving improved efficiency and reducing carbon footprints.
Smart meters provide a remote read-out of consumption information in real time and allow utilities and consumers to better track usage. ASICs are used to measure voltage and current conditions, interface the meter to the outside environment by a variety of RF communications standards with the added functionality of GPS to allow for time and location measurements.
Linear and Rotary Encoders
Many industrial control systems use linear and rotary encoders. A linear encoder is a sensor, transducer or readhead paired with a scale that encodes position. The sensor reads the scale in order to convert the encoded position into an analogue or digital signal, which can then be decoded into position by a digital readout (DRO) or motion controller. The encoder can be either absolute or incremental (relative), and motion can be determined by change in position over time. Linear encoders, which may include optical, magnetic, inductive, capacitive and eddy current technologies, are used in metrology instruments, motion systems and high precision machining tools ranging from digital callipers and coordinate measuring machines to stages, CNC mills, manufacturing gantry tables and semiconductor steppers.
A rotary encoder, also called a shaft encoder, is an electro-mechanical device that converts the angular position or motion of a shaft or axle to an analogue or digital signal. As with linear, rotary encoders are either absolute or incremental relative. The output of absolute encoders indicates the current position of the shaft, making them angle transducers, whilst the output of incremental encoders provides information about the motion of the shaft, which is typically further processed elsewhere into information such as speed, distance, and position. Rotary encoders are used in many applications that require precise shaft rotation: industrial controls, robotics, special purpose photographic lenses, computer input devices (such as optomechanical mice and trackballs), controlled stress rheometers, and rotating radar platforms.
Today, it is possible to integrate substantial functionality into an ASIC, especially at the smaller 22nm and even 14nm CMOS geometries. However, these geometries are best suited for predominately digital designs so, for industrial mixed signal designs, the favoured geometries tend to be in the 350nm – 110nm range. These can still achieve a high level of digital integration, including DSPs and microprocessors, as well as offering the flexibility for the analogue part of the design. Power devices and high temperature solutions are also possible, opening up new application areas for integration onto the ASIC.
Communications & Flexibility
In some cases, the solution is for a platform approach, where one ASIC will be used in different scenarios within a product family, and it will support many sensor interfaces and communication
ports. This type of architecture can perform information processing local to the sensor and report back to the main controller. Being generic allows it to interface with many different sensor types, and the controller communications can be wired or wireless. It can also be programmed, so has some flexibility.
ASIC Design Team
Whilst mixed signal ASICs are changing the face of industrial controllers, it is important to note that an ASIC design is not simply a question of implementing an existing PCB design.
Different techniques are needed for ASIC design, and good system knowledge and experienced designers are essential for a successful project. In many cases, this comes from an amalgamation of the ASIC design team and the system designers, so it is crucial to build this relationship early in a project.
Design NRE is calculated on an internal rate over the estimated project. There are often complex calculations for this and some really quite simple. At the end of the day it will be the price for getting the job done, it will include any specialist expertise, tools or equipment required to get the job done, such as testing, digital, RF, etc. It will be based on a project plan. The questions you may want to ask, are, is this fixed price and what contingency is in here? It is important to conduct a risk assessment at this juncture…a slipping specification guarantees re-spins as does trying to break the laws of physics!
Making the Right Choices
For the design NRE, on the surface this seems relatively straight forward. However, when undertaking any ASIC design it is critical to make sure you have the appropriate skill sets in place.
This may sound obvious but it is often overlooked. The final goal of any ASIC project is to have a fully qualified, working device at the minimum cost. So selecting the right resources will have a major impact on this. A digital design requires different skills to an analogue design, so a partner with the correct skill sets should be selected. Also it is not just the design that has to be considered at this stage, the production and testing of the device should be included. The wrong process selection could lead to escalating cost. So any partner should be able to show the whole ASIC process and not just the design.
The design process also requires careful portioning expertise, (i.e. what is going to be implemented onto the ASIC), as it is possible to put many things onto the silicon.
The Factors to be Considered
When undertaking an ASIC design, many factors have to be considered. Some of these are obvious where as others may be more subtle. For example, the design NRE, obviously the more complex, the longer it will take to design, layout and test. It will also increase the size of the device and test time during production. It will also affect the yield as a larger design has a higher probability wafer imperfections and localised defects will cause chip failures. So the effective partitioning of a design is critical to its success. However there can substantial savings by incorporating a microprocessor into the ASIC, and such decisions have to be balanced with any licensing fees. Look for a mature development processes, that are well documented and examples of previous designs that are representative to your requirements. Included in these should be a detailed ASIC specification and test strategy for qualification and production.
Careful on Costs
Any IP NRE can add significant costs to this stage, (also potentially to the unit price if a royalty based model is employed), but could save design time and reduce BOM costs. When considering using IP, the designer needs to consider if the IP has been to silicon in the selected process, also for some IP, what development environment is available. This is critical for processor IP as there will be a need to develop code and potentially program the device. Also any qualification requirements need to be taken into account. For example, has an RF block been qualified and certified previously. Any specialist standard requirements have to be accounted for, (e.g. ISO 262262), at the design and IP selection stage.
There are also many fabrication processes available, and again, the selected process should be carefully matched to the design elements and ASIC performance criterion. As discussed previously, as the process geometry decreases, the mask costs rise quickly and certain functionality becomes more difficult, so the selection of a smaller geometry has to be carefully considered and it is not always the case that to select the latest process available is the best choice.
Whilst the design of the device is critical, it is not just the functionality that has to be considered, as the device has to be manufactured and tested, and decisions at the design phase will impact the yields and hence cost of the development and parts. Yields can be influenced by the process selection so again a partner who has detailed knowledge and access to a selection of fabrication plants and processes will be able to offer the better choices. The target yield should be defined at the start of the project. The product life cycle is another critical factor, in that if the device is to be available for many years, the right process has to be selected to enable this, so an obsolescence assurance policy is desirable.
For some industries there are strict constraints on field failure rates, which have to be mitigated by using part average testing techniques to maintain rates of less than one part per million. This involves statistical analysis of parts at the limits of the specification and rejection of part that have a higher probability of field failure, even though they are within the specified design limits. This can be achieved with adjacent failing part reject techniques, and additional stress testing to identify likely field failures.
The Major Decisions in ASIC Development
So again the right partner selection is critical if you are targeting these markets. Essentially a partner with all the relevant expertise of the ASIC development and supply chain is required. This has to include logistic support and fast reaction times should a problem arise in production, so tight control over testing and an active failure analysis strategy. This will allow you to be able to treat the ASIC as an off the shelf part, confident in the knowledge that should a problem arise it will be addressed in a quick and professional manner. If you are targeting some markets, this is a requirement not a nice to have. The skills required are detailed in the table shown below.
It can be readily seen that the option of using a full service from a company like Swindon Silicon Systems can fully embrace every aspect of design, development and delivery.
Working in Partnership
There are many issues in an ASIC development and production project which impact of the choice of partner and the costing and competitiveness of the final outcome. That is why, at
Swindon, we have evolved a mutual approach with the aim of working in partnership with our customers to achieve the right balance of decisions and choices in terms of development and
The chart below details the approach form our initial contact through to the delivered product.
The Advantages of ASICs
ASIC’s have many benefits and can give you significant market advantage and cost savings, so before you dismiss this option it is worth checking if it is appropriate for you. The costs and benefits may be a pleasant surprise. Swindon has all the experience and expertise to guide you through the many choices you have. With mature development processes, and excellent relationships with the fabrication plants, fully qualified automotive capability, with 100% inhouse testing and over 300 hundred successful projects. When it comes to ASICs, Swindon has all the answers.