Latest news

Altera and Eutecus Single-chip, FPGA-based Solutions “See” and Provide Intelligent Vision for Smart Cities

Altera Corporation and strategic IP partner Eutecus are announcing the availability of the ReCo™-Pro Multi-channel High Definition (HD) Video Analytics platform based on Eutecus MVE™ video and fusion analytics technology and Altera’s Cyclone® V SoC and Enpirion® PowerSoC devices.
Available from Eutecus, the ReCo™ platform has been chosen by Sensity Systems as the foundation for adding intelligent vision processing to its high-speed, Light Sensory Network (LSN), which is currently being installed in several US metropolitan ... Read more Read more

Home Automation with Raspberry Pi and an Embedded Touchscreen Display

Remembering which light switch controls each light in your home can be difficult sometimes, especially when three-way switches and more complicated setups are thrown in. Compound that complexity with the amazing versatility of home automation systems, and simple switches start limiting the potential to fully control the lighting environment.

By Tyler Crumpton, Makers Local 256, Huntsville, AL USA

To fight this growing complexity, I decided that I wanted a “smart” light switch that would let me know what lights in the house are currently on and allow me to turn any light on or off without having to figure out what switch goes where. It needed to be custom-fit to my home, yet flexible enough to be useful if the number of lights are changed, lamps are added, or if someone else wants their own switch in their own home.

Acrylic floor plan with ELI screen and Raspberry Pi
There were not any solutions that fit my needs, at a low enough cost, so I turned to my trusty Raspberry Pi (RPi) single-board computer to help me make my own smart switch.
Since RPi has USB ports and general-purpose input/output (GPIO) pins, it could be used to control ZWave, Zigbee, or Wi-Fi smart lights using a USB dongle or to control some LED indicators for debugging and testing without having to run around the house to check if lights are turning on or off.
Since having a simple LED home lighting “simulator” would be extremely helpful in developing a smart switch application, I went down to the local makerspace, Makers Local 256 in Huntsville, AL and started creating a vector image of the house floorplan. This floorplan was then laser cut from a few sheets of acrylic plastic which were glued together afterward.
LED lights were then glued to the back of the floorplan, one LED in each room, and wired up to the GPIO pins on the Raspberry Pi. Using a small Python script, I was able to turn each room’s LED on or off with a command.
With the simulator complete, I began working on the switch interface. I decided that the easiest and quickest way to turn a room’s lights on was to just tap on a room in a floorplan, since touchscreens have become second-nature to most of us. I wanted the floorplan interface to be available from a computer, smartphone, or even a dedicated touchscreen device mounted on a wall where a light switch would normally be.
Using the Tornado Python web framework and a little bit of Java script and HTML, I wrote a tiny webserver that displayed the floorplan of the house and let you click on each room to change the light. The room would glow when the light was turned on so that I could tell what the state of each room’s lights was in. The server sent a command to the simulator each time a light was changed, so each tap on a room lit up or turned off the corresponding LED on the simulator.

Completed simulation with LED lights
The web application worked perfectly on a computer or smartphone, but I still needed a dedicated replacement for the old light switches on the walls. Since the Raspberry Pi that ran the web server has an HDMI output and a spare USB port, I decided to use the ELI70-CR 7.0” embedded color touchscreen from Future Designs, Inc. to display the web application.
The ELI, which stands for Easy LCD Interface, is an off the shelf embedded touch screen LCD solution for use with an SBC like the RPi. The screen required no extra configuration other than switching to portrait mode and I set the Raspberry Pi to jump straight into the floorplan controller whenever it boots.

7.0” Easy LCD Interface (ELI)

I have not yet decided on a wireless technology for controlling lights in my house, so the next step is to install ZWave, Zigbee, or Wi-Fi lights or light controllers. There are many projects for ZWave that have Python mappings, so making the web server control those devices via a ZWave USB dongle would only require a few lines of code.
Changing the floorplan for use with someone else’s house is as simple as drawing up the plan and defining the rooms. There are so many options that can be added, including dimming control, color control with the Phillips Hue, and energy conservation by scheduling lights to turn off at certain times.
Overall, I was able to create a custom, personal light controller for my home using only a Raspberry Pi, the FDI ELI70-CR touchscreen, and a USB dongle for ZWave, Zigbee, or Wi-Fi.
It took less than a day to get most of it working, so expanding upon the initial design should be quick and fun, and hopefully make my home just a little bit easier to use and a whole lot cooler.

Find more information:

ELI (Easy LCD Interface) from FDI:

Makers Local 256:
Read more

Sentinel – always on guard

Kontron Box-PC makes Wattics’ energy analytics system maintenance-free.
Although reducing energy use is on everyone’s agenda, organizations face difficulties in identifying saving opportunities. Wattics Sentinel is a self-learning energy analytics system providing the solution: The system converts all the complex meter readings into insightful information, and notifies customers with clear recommendations. In the past, the local intelligent platform powering up Sentinel produced considerable maintenance costs. This is why Wattics has spruced up its solution and now uses German manufactured Kontron Box-PCs in ‘wartungsfrei’ (maintenance-free) quality.

Author: Ingrid Einsiedler, Marketing Manager, Kontron

With increasing energy costs, companies need to significantly reduce their energy consumption and find solutions to identify energy waste and inefficiencies. Conventional commercial solutions will often only deliver dashboard tools showing ‘big data’ from all the different measurement devices such as utility meters, building management systems and other electricity, gas and water meters. Unfortunately, in the majority of cases the time and/or knowledge of how to process all this information in order to identify improvement opportunities are lacking. Companies consequently desire intelligent solutions to support them in this analytics challenge more or less automatically. With its Sentinel energy analytics solution, the Irish company Wattics offers the solution: built around intelligent self-learning algorithms the system automatically identifies energy waste and cost-optimization opportunities for the end customers. It also features an innovative monitoring gateway for cloud services delivering round-the-clock, instant information – accessible from anywhere – to accredited staff.

Self learning software solution

The Sentinel energy analytics solution works with an innovative software engine that – immediately after installation – automatically begins to learn from collected meter readings to identify energy use patterns. The solution thereby keeps watch over energy usage across all monitored meters from the main energy distribution board to equipment

The Wattics Sentinel notifications are forwarded through the Wattics Messenger Dashboard, to an email or mobile phone depending on urgency and preferred settings
level. It constantly improves its knowledge about the operation of appliances and grid areas, identifying correlations between their use models, and characterizing good and bad energy use. After the initial learning phase, the system starts notifying area managers and operators according to settings. Crucial energy issues are displayed via clear text messages which provide detailed and easy understandable information and recommendations which can be displayed on a range of different devices such as smartphones, tablets or notebooks and desktop PCs.

Compatible to any existing infrastructure

The information provided by these messages is not data-centered but alert- and activity-oriented which prompts users to focus on the implementation of the recommended actions. This ultimately culminates in increased power savings which are achieved more efficiently. The messages and recommendations the system generates are not static, but they are continuously improved and adopted according to best practice comparisons, optimal usage values, standard figures and comparable installations. A further benefit of the Sentinel energy analytics solution is its compatibility to any common meter from site to equipment level. The investment is therefore future-proof and its interoperability helps customers to remain independent from meter manufacturers. Another benefit of the Wattics solution is the fact that a local hardware investment in such an analytics system is limited to a single monitoring gateway.

The central energy monitoring gateway

Wattics will deploy one enterprise-class Sentinel energy monitoring gateway per customer site requiring energy analytics, with the benefit that no additional communication hardware is required. This gateway is used to retrieve the meter readings through serial and/or Ethernet communication and to secure data communication towards the cloud to the system backend, hosted by Wattics. Additionally, the gateway manages pre-processing the collected data to analyze variations on power measurements. It also acts as a local back-up for easy recovery in case of an internet failure as well as a data compressor to reduce storage demands in the cloud. The energy monitoring gateway is connected to the internet via the customers’ existing wired LAN or a wireless 3G router.

Features of the Wattics Sentinel energy monitoring gateway

Run the Wattics user interface to configure metering deployment (discover meters, associate circuits to meters, configure meter and current transformer settings and set up internet)
Data Collection
Manage reliable retrieval of meter readings through serial and LAN communication
Data Back-up
Local back-up in the event of internet failure recovery
Data Pre-processing
Analyze variations on power measurements to identify initial list of energy events of interest
Data Compression
Provide compressed data and reduce cloud data storage through data pre-processing service
Data Secure Communication
Handle secure communication with Wattics cloud servers via API

The Wattics Sentinel energy monitoring gateway based on Kontron’s Box-PC KBox-A101 is the central ‘edge node’ at the customer’s site. It is connected to both the local energy grid as well as to the Sentinel grid built on modern IoT technologies.

Managing costs and reliability

The Sentinel energy monitoring gateway is instrumental with both the overall availability of the energy analytics solution and the cost efficiency of the installation at the customers’ factories, shop-floors and buildings. As such, the gateway needs to be highly reliable. In installations with a required long life like in the case with Wattics solutions, maintenance requirements are particularly crucial. This is due to the fact that distributed solutions often increase costs, which can easily match or even exceed the investment cost for the entire gateway system.
Even the smallest of failures can call for maintenance work which, ideally, should not occur at all. If, for example, the gateway operating system would be causing problems or the system would crash for whichever reason, this would at least require a manual reboot. Plus, in order to identify and investigate the kind of failure, the client would have to run a number of checks. Based on the results, an on-site visit might be required and the hardware may need to be serviced, causing an increased workload as well as additional costs. In case of a distributed usage model with nation- or even worldwide installations, avoiding these failures is mandatory in the interest of economic operation and customer satisfaction.
With the Kontron Box-PC KBox-A101, Wattics finally found a platform to fit the bill. It offers German manufactured ‘wartungsfrei’ (maintenance-free) quality so that Wattics can now cut maintenance costs and offer a long-life solution to their customers.

Maintenance-free system designs

To achieve a maintenance-free system, industrial computers have to be built without rotating components. Rotating components wear out, are vulnerable to shock and vibration, and have to be exchanged regularly in long-term installations. Maintenance-free systems offer a fanless design in combination with the latest energy efficient processor technology. With regards to data and OS storage, maintenance-free systems rely on durable and fast Solid State Drives, ruling out mechanical failure. Another critical factor is the power supply. Some systems which Wattics previously installed experienced failures after just two years of operation. High quality power supplies that are durable and available for the entire lifecycle are crucial to maintenance-free designs. What’s more, this type of high quality power supply can also often handle power failures of several milliseconds. This valuable additional feature prevents system reset and consequently serves to improve system availability.
Maintenance-free systems do not require classic button cell batteries to continuously power the BIOS or the EFI memory and the internal clock system which would have to be replaced every two to three years. Instead they use wear-free double layer capacitors – so-called gold caps – which do not require replacement and thereby maintenance. In addition to these inner values, maintenance-free systems also feature a rugged as well as dust-protected casing. This helps to further increase the reliability of the whole installation even in harsh environments, as can often be found on the shop floor or in other industrial environments.
All these features contribute to a high Mean Time Between Failure (MTBF) and consequently to the maintenance-free design of the application, provided that the MTBF is in line with the life cycle expectations of the application. The Kontron Box-PC used for the Wattics Sentinel gateway has an MTBF of 70.000 hours. That corresponds to 24/7 operation of up to eight years or in single shift operation up to 24 years. From a technical perspective, during this period the system does not require any maintenance.

For Wattics, the high MTBF in combination with fair pricing and the long-term availability of the Kontron Box-PCs of at least 5 years delivered the winning arguments for us to choose an industrial computer out of Kontron’s new ‘wartungsfrei’ range. Further to this, the global availability of Kontron technologies and services fits perfectly into our long-term strategy. We’re convinced that with the Kontron system we’ve found a really sustainable solution,Anthony Schoofs, CTO at Wattics, comments ■

Additional resources:
The Kontron Box-PC KBox-A101 product website.
The press release about the Kontron Box-PC KBox-A101 launch:
More about Wattics and Sentinel can be found at the Wattics website:

Contact Details
Ingrid Einsiedler, Dipl. Wirtsch. Ing.
Marketing Manager, Kontron AG / Site Kaufbeuren
Sudetenstrasse 7
87600 Kaufbeuren, Germany

Wattics contact details:
Anthony Schoofs, CTO
Wattics Ltd
31/33 The Triangle, Ranelagh
Dublin 6, Ireland
Read more

Wearable health devices and power management

Inspired by element14’s Sudden Impact design challenge, this is the third in a series of exclusive blog posts for MDT that explores the challenges of creating wearable medical devices.

Author: Christian DeFeo, eSupplier and Innovation Manager, Newark element14

As our Sudden Impact design challenge comes to a close, we take a look back at the last stage of the design process and see how our twelve finalists’ prototypes coped when put to the test. While trialling their wearable health solutions across various terrains, our competitors unanimously faced one key challenge – power management.
The process of collecting and analysing data over long periods of time has had an inherent impact on energy consumption in our competitors’ sport-related designs. As such, the finalists have been trialling both new and established power management platforms. In this blog, we take a look at how these are being used to optimise their designs.

Wireless charging

One Sudden Impact competitor, Douglas Wong, has been looking at wireless charging as a long-term solution to powering his helmet-mounted trauma monitor for hockey players. To ensure his device remains user-friendly and weather resistant, Douglas’s approach was to embed a Lithium polymer battery into the helmet, enabling it to be charged wirelessly using a Qi charger.
Qi charging is a global standard developed by Wireless Power Consortium that enables any device with a compatible battery to be charged from a wireless pad, using induction transfer. This is a standard that many smartphones already adhere to, and its practicality means that it is beginning to filter into the wearable health industry too.
From a usability point of view, Qi charging provides a simple solution for hockey team managers. By using handful of Qi charging pads, the entire team’s helmets could be fully charged before a game which would consequently ensure the safety of every player on the pitch.
However, with that said, at this stage it is likely that designs such as Douglas’ may still be too much for a QI pad to handle. While the concept has been proved, for this to be a feasible long-term solution, the Qi system needs to be implemented on a significantly larger scale to enable wireless charging to become second nature.

Bluetooth 4.0 and BLE

Another Sudden Impact competitor, Hendrik Lipka, has been trialling various Bluetooth protocols for his helmet-mounted impact and heart-rate monitors, aimed at skiers and footballers. Although the terms Bluetooth 4.0 and Bluetooth Low Energy are often treated interchangably, during his design process, Hendrik discovered just how different the two standards are.
Bluetooth 4.0 is a relatively new type of wireless technology, offering considerably lower power consumption versus previous standards. It is a combination of three different protocols: Bluetooth Classic, Bluetooth High Speed and Bluetooth Low Energy (BLE). With the exception of their data transmission processes, Bluetooth Classic and Bluetooth High Speed are relatively similar. BLE, however, is designed for extremely low power devices and works best for short lived, low-data transmissions.
As such, during the design process, Hendrik found that BLE is particularly useful for the transmission of real-time information. Hendrik’s heart rate sensor could therefore capture an athlete’s heartbeat as a ‘current state’, alongside the minimum and maximum values within a specific time frame. As such, BLE proved itself to be an appropriate protocol for athletes that want to use the heart-monitor as a safety measure and be notified of irregular heart rates. However, for continual monitoring, BLE is not ideal and sheds light on the functionality vs. longevity battle that engineers are constantly grappling with.
Ultimately, present day power management solutions still have some way to go before they are totally suitable for use in the wearable health market. The methods that our finalists have been testing highlight the need for more data-intensive technologies that can relay large quantities of data for long periods of time.
In our next and final blog, we look at the final steps our Sudden Impact finalists have taken to elevate their design from an idea, to a viable and usable device ■
Read more

Global Support for Automotive Manufacturers and Suppliers

Rutronik Automotive Business Unit
With its newly founded Automotive Business Unit, the distributor Rutronik promises to provide focused support to manufacturers and suppliers on a global level. Uwe Rahn, Senior Manager and head of the Automotive Business Unit, explains this in detail in the interview below.

Uwe Rahn, Senior Manager and head of the Automotive Business Unit,
Rutronik Elektronische Bauelemente GmbH

What can customers expect from the new Automotive Business Unit?

Uwe Rahn: The customer will receive technical and commercial support at eye level from a team of experts with many years of experience in the automotive industry. We will be working in close cooperation and coordination with specially selected manufacturers. In addition, we will also be providing support to our European customers in their efforts to gain access to the American and Asian markets.

Which services in particular does this include?

Integrating electronics into automobiles on a mass scale due to increased networking, focusing a greater amount of attention on comfort, infotainment and assistance systems, not to mention e-mobility technology, has created the need for components, which originally were not even developed for automotive use. Any developer of new vehicles now faces huge challenges as a result of offering Internet connectivity in a vehicle and the more-or-less indirect communication with safety relevant systems, among other things. With our experience in both the automotive and electronics industries and our close ties to manufacturers, we can offer valuable support in these areas, allowing our customers to concentrate fully on their own applications.
Of key importance in our global business are the central information technologies that are available around the world. One important component here is our ability to centrally handle all PCNs and EOL messages in one database, which then makes them available worldwide to every customer and manufacturing service provider in the accustomed data quality. These IT systems are now being expanded step-by-step, by adding such information as standards and certifications, including AECQ-100 or 200, PPAP and APQP (Advance Product Quality Planning), Failure Mode and Effects Analysis (FMEA) documents, and even VDA compliant process descriptions.
Process support is yet another important aspect. Many Tier 2 suppliers now have an even more stringent requirements profile. We can provide them with the support they need in developing and implementing processes similar to the support we provide to Tier 1 suppliers.
The third large component here is our ability to support our international customers in the North American and Asian markets.
To do so, we will make the focus of the business unit a global one by working hand-in-hand with the regional RUTRONIK branches and our manufacturers locally.

Which manufacturers do you cooperate with in the new business unit?

We will be gradually expanding this current network over the next few years, yet our focus will always be on quality, not quantity. Our selected partners will have to meet a range of criteria, which we consider to be absolutely necessary for our customers, primarily an innovative portfolio of automotive products and a global presence.

What does this cooperation with these partners look like?

The members in our unit deal very closely with the contacts among the group of manufacturers. They receive special training on working with the derivatives at a very early stage. On the other hand, we provide our manufacturers with information on the most current challenges that our customers face. We are the pulse of the market, so to speak. For several years now, we have been offering the Tech Days event, where a selected group of customers and our special manufacturers come together for a concentrated exchange of ideas. Typically, this is followed up by a customer-specific Expert Design Workshop with one manufacturer in particular. Here, a specific customer application and its system architecture are discussed.

What kind of products are currently being focused on?

In addition to the new semi-conductors in the micro-controller and MOSFET fields that are currently being developed for automotive applications specifically, these are, for example, MLCCs, which are ceramic multifaceted capacitors used in safety relevant applications. AVX has developed its own Automotive Plus series, which complies with the AEC-Q200 standard. These MLCCs are extremely reliable and resilient against mechanical and thermal influences.
Some products can even be used at temperatures up to 250°C. The new capacitors feature an extremely high current carrying capacity and high insulation resistance coupled with low ESR / ESL. They were developed for highly sophisticated applications, such as high pulse current switchings, aviation and aerospace technology as well as for use in hybrid automobiles.
Another example from the electromechanical range of products are the new crimp contacts, which we offer from JAE. They make it possible to connect any kind of wire to a circuit board quickly, easily and cost-effectively, yet while creating a connection that cannot be easily detached, featuring a high degree of electrical and mechanical reliability. In automotive applications, the connection is approved first. Not until then is the housing developed. Omitting this housing makes it possible to save costs considerably as well as a great amount of space.

Which applications do you feel currently have the greatest need for consulting?

These would certainly be such current trends as: e-mobility, car-to-car communication and networking within the vehicle, including infotainment, advanced driver assistance systems, so-called ADAS, and even autonomous driving in its various stages of development from driver only to assisted, partially assisted and highly to fully automatic. Moreover, with the constantly increasing safety requirements as per ISO 26262, which cover on-board applications in various voltage ranges from 12, 24, 48 and up to 450 volts for hybrid systems, there is a huge range of topics that require in-depth consultation.
LED lighting as well is still a huge topic. Today's state-of-the-art applications for LED daytime driving lights have a power consumption of only about 15 watts - this while at the same time offering the ability to be more visible to other road users. This equates to savings of around 0.2 liters of gas per 100 kilometers, which equals 4 grams fewer CO2 emissions per driven kilometer.
The next phase of development in this field of application will be realized with OLEDs as of 2017. Where in the past the use of OLED technology was limited to high-resolution displays, for example, in the future they will also be used in automotive vehicles as a tail light or a turn signal. The main reason for their use in the future is that they require very little space and can also be designed as convex or concave light sources. However, OLEDs will also be a modernizing feature for the interior. Currently, they are being developed as ambient solutions in areas such as the footwell, the ceiling panel, the dashboard and door lighting in various design options.
Finally, the entire vehicle design concept is currently undergoing significant change as in hardly any other industry. These developments shift from mechanical to electronic innovations. This is why there is so much uncertainty and the need for consulting is extremely high right now. At the same time, European suppliers are under increased competitive pressure from new suppliers, especially from BRIC countries such as Brazil, Russia, India, China and South Africa. Here, we feel we are well positioned to offer our customers the support that they need.

What distinguishes the Automotive Business Unit for this purpose?

First of all our many years of experience. The automotive segment takes up more than 45 percent of our business at Rutronik and has thus always been one of the most important business segments. This means: We know what our customers need. We speak their language, know their processes, requirements and rules. And for many years we have also had close contacts with the most important manufacturers in this innovative sector.

This also applies to the range of products you have offered in the past. So why do you need a separate Business Unit?

Our aim is to bundle the specific support, expertise and synergies that the various manufacturers and their products can offer and make them available to our customers on a global level.

Would you like to briefly introduce the Business Unit?

Yes, I'd be happy to. Our team is made up of experienced special FAEs for automotive applications, who are supported by their colleagues from outside sales, inside sales and back office administrators. In addition, the automotive specialists act as an interface to the colleagues from the product sectors, such as power train, micro-controllers, passive or e-mechanical products. The global unit is managed by myself in the Ispringen headquarters. My business partner Lutz Henkel provides support to the North American market within the unit. We both have well over 25 years of experience in a number of positions at Tier 1 suppliers and distributors. Our colleagues in our branch offices around the world provide support to our customers locally and coordinate with us closely ■


Read more

Internet of Things, Hype or Hypertension?

“Internet of Things” was one of the biggest buzz words of 2014 however were the interconnected technologies it tries to describe really novel, or rather the evolution and expansion into new areas of existing programmes such as Smart Metering or Grid, Home Automation, Industry 4.0 and Intelligent Automotive Systems.
So how are the connected innovations these prior programmes initiated and proved finding their way into other market areas and what are the benefits and challenges to be overcome?
As we move into the new year of 2015 will the term “IOT” become stale and fade and will the progression of interconnection between embedded devices continue, without the headaches associated with the issues to realising tangible business opportunities being faced.

Author: Joachim Hüpper, Industrial & Communications Business Unit, Renesas Electronics Europe GmbH

How is IoT changing the world

There are one billion people on the internet today and Gartner predicts 26 billion devices will be online by 2020, a 30 fold increase from 2009. This presents a $300Billion incremental business opportunity to the ecosystem of suppliers to the embedded market and services providers for the connectivity and data management. There are many areas of where IoT connected embedded devices can bring real benefit and are currently being developed & deployed, some examples of these being:

Smart Meter deployments have on going in many European countries with the EU mandate to have 80% of European homes using Smart Electricity and Gas meters by 2020. Most of these programs have a HAN (Home Area wireless Network) and WAN (Wide Area Network back to the supplier via the internet). This brings the possibility of meter connection to smart thermostat, boiler and radiator valves a close reality. British Gas in the UK have already released the HiVE app for controlling heating via a smart phone, with boiler and thermostat connected via Wi-Fi to router.
These is of course also the much publicized “Google Nest” connected learning thermostat which is being enhanced by the Google “works with Nest” programme , this would enable for example a Phillips Hue lighting system to flash the lights if the “Nest protest” sensor detects carbon monoxide.
The features of these early leaders may well become standard in years to come, providing greater control and energy saving which will be great for consumers and the environment.
Also in the built environment, printers have had the feature to prompt users for online ink cartridge replacement for many years. Home Appliances and AV equipment may extend this further being able to report fault codes back to the manufacturer or provider, initiating service calls or spares and consumables to be ordered. No more digging around in the back of the user guide for the fault finding table or searching for a spare part supplier.

The next generation of interconnected manufacturing equipment has become known as “Industry 4.0” where intelligent factories, machines and products communicate with each other, cooperatively driving production. Companies such as Siemens are in advanced stages of research and development of these systems. Enhancing this further smart tags will provide connectivity of the ingredients required for the manufacturing process, giving asset management & materials tracing, reducing errors, providing improvements to “Just in Time” processes and reducing theft.

Connected sensors to detect environmental changes and report early warning for earthquake, avalanche, eruption & tsunami have existed for some years. Now there is the start of the installation of sensors and beacons in major towns and cities. Trials are already under at many sites including San Francisco airport and the city of Reading UK to aid the visually impaired navigate their environment via beacons that send audio announcements of where they are and what is close by to a headset. At San Francisco Airport when users walk past one of the 500 transmitter beacons, their iOS device will announce nearby points of interest; they can find flight gates, ATMs, information desks and power outlets without asking for help.

This could be projected further into the future with self-driving cars avoiding congestion and intelligently rerouting for optimal flow by receiving information from beacons en-route connected to a central traffic management system.

With products like smart watches and the creation of sensor beacons, advertisers have started to realize the massive potential of all this data being collected, and how it can be used to reach consumers. The potential of target adverts being pushed to consumers smart devices as they pass by or browse within a store are not far off and have already been presented in the 2002 Tom Cruise film Minor Minority Report.

In a similar vain to manufacturing in the industrial sector, retail could also benefit greatly from internet connected products. Tesco in the UK generated almost 30,000 tons of food waste in the first six months of 2013. Industry-wide, 68 percent of salad sold in bags are thrown away . IoT offers the prospect of better tracking and management of food items to reduce wastage.

The recent explosion of home healthcare products such as blood pressure & heart rate, blood glucose & body mass index, lend themselves perfectly for evolution to connected devices which send the collected data to predict trends and send to services for analysis.
At the Consumer Electronics show this year Connected Health was centre stage.
Monitoring of body signals and activity using wristbands such as FitBit is now widely accepted and the launch of Apple Healthkit in iOS 8 will initiate many more connected monitors.

What are the new challenges that this evolution brings?

So behind the hype there is some reality today but for each of these application areas to become mainstream there are still many challenges. Five of these we will now consider.

For distributed connectivity wireless is the most practical solution but there is massive fragmentation. Cellular technologies are well established but can’t provide a solution for battery or energy harvested devices such as sensors or beacons. There are no shortage of wireless standards in fact a mind numbing array: Wi-Fi, Wireless-Mbus, Bluetooth and IEE 802.15.4. Each of these may be a good candidate but many have sub standards and frequency or software stack options such as ZigBee, 6LowPAN,

Thread for ‘802.15.4, each with different stack versions and profiles, causing concerns for interoperability.

The Industrial Internet Consortium was founded to help resolve these problems and has a following of greater than 100 member companies, but results are yet to come.
Wireless solutions also need careful antenna design and optimisation plus certification to ensure compliance with the chosen wireless standard, both of which are complex fields and require specialist skills.

Security is probably the greatest headache to most. Many IoT application areas could result in some financial transaction, so a lack of tight security would leave systems open to fraud and any linked to infrastructure could be a target for terrorism, for example energy or traffic management.
Many embedded development companies do not have the skill or expertise in house to ensure robust encryption and authentication so will need to reach out to external experts. For the end consumer, data protection will become a big topic, with remote sensors and beacons tracking our every movement and activity, ensuring that this data is kept private and the consumer feels they are in control, will be key for acceptance. A number of much simpler smart metering trials were aborted due to consumer pressure on this issue.

Big Data
A further challenge for the evolution of IoT will be what to do with all the data that results from the connected sensors. A years’ worth of UK smart metering data of ½ hourly gas and electric meter reads is approximately 3Tera bytes , which come from essentially just two connected sensors. Therefore professionally managed IT web server infrastructure will be required for collection, aggregation, analysis and presentation of the potentially massive amount of data created from IoT systems.
As well as collection data there is also the enablement of download or pushing data to the remotely connected devices. Careful design and testing will be required to support features such as remote firmware code module update, feature addition or deployment of apps in interpreted languages such as Java, python or other languages, including appropriate security features to ensure only authenticated updates are accepted.

Power source limitations
The next technical issue to be considered will be the power source for the end sensors. As most end devices in an IoT environment will not be mains supplied, consumers will reject them if constant battery replacement is required. As the quantity of end devices increases so must the battery life of each to avoid consumer inconvenience. If you consider that 24 devices with a battery like of 2 years each has the same replacement rate as 1 device with a battery life of a month. So efficient low power component choice, system design and battery less techniques such as energy harvesting will need to be embraced.

Use cases to be translated to business cases
Finally, ideas need to be translated into money. There will no doubt be many IoT products launched, but it will be ones with well thought out business case that will thrive. Interestingly as we will consider next, IoT will bring more possibilities for a positive revenue stream than simply the value of the physical device as has been in the past.

So how will this benefit embedded solution providers?

For embedded solution providers, the evolution of an internet connected world which IoT promises, opens up many opportunities and will probably be a disrupting influence in a number of application areas where slow moving dominant players may be displaced or lose market share by more dynamic innovative new entrants. Some examples of these opportunities are:

Differentiation & Innovation
Platforms and interconnect with sensors and actuators can allow differentiation from competitors or new entry to disrupt market areas currently serviced by others. A good example here is integrated home environmental control, where a supplier of security systems today could differentiate themselves by enabling security presence sensors to interact with lighting control or heating system providing energy management or vulnerable person monitoring. “Alertme” in the UK is a good example of this.

Offer new services
For consumers, new services that can increase customer satisfaction or create new revenue streams will appear. Data services, for collection, analysis and presentation of the information from IoT sensors and provide control of connected actuators. These services may be provided directly to the consumer or via other parties that produce “mash up’s” with added value of data from different sources. Think how google maps has been combined with house prices from property sales portals to produce heat maps of property values for an idea of what may come next.
For the developers of IoT products, sourcing proven platforms of software and hardware will be a key enabler, especially as many of the technologies required such as wireless and security are beyond the capabilities of existing equipment manufacturers and certainly start-ups. Communications modules are an easy option for quick addition of wireless technologies particularly whilst de-facto standards for IoT are very much in flux. Finally consultancy services will be very much in demand especially for security and communications technologies.

Increase or add value
Electronic products prices are continually reducing, whether its security systems, appliances or semiconductors. With the advent of IoT adding new innovative features to products that differentiate, will allow product prices to be maintained, whilst offering new services that make use of the collated data or provide control of connected actuators, allowing new revenue streams to be created.
Going forward companies have already identified that the model of revenue purely from hardware is being replaced by profit and value from solutions, software & services.

In this article we have shown that though there is a lot of hype around the term Internet of Things there is much reality today with immense opportunities for the future. However to fully realise the opportunities of internet connected devices will require new IT infrastructure, embedded technologies, software, standards & certifications that end equipment designers and manufacturers today have limited experience of.
A way forward to enable developers is for proven platforms combining wireless, security, and low power operation along with IT data infrastructure that are “production ready” with high quality software stacks, integrated, tested and qualified to required standards ■
Read more

Maxim Integrated Demonstrated Highly Integrated Analog Solutions at Embedded World 2015

Maxim Integrated Products, Inc. demonstrated highly integrated analog solutions for embedded applications at the Embedded World
2015 Exhibition and Conference in Nuremberg, Germany (February 24–26, 2015). Organized in three demo areas for Industrial Power, Industrial Interface, and Signal Chain, Maxim’s solutions showed systems engineers how to simplify designs and get to market faster.

Industrial Power

Step-down DC-DC converters eliminate external components and reduce total cost. A “demo in a box” features Maxim’s Himalaya series of highly efficient, 4.5V to 60V, synchronous DC-DC buck regulators from 25mA to 3.5A: MAX17552, MAX15062, and MAX17501 / MAX17502 / MAX17503 / MAX17504 / MAX17505.
Cooler, smaller, and simpler DC-DC step-down power modules reduce design complexity, manufacturing risks, and time to market. The Himalaya series pin-to-pin-compatible power modules, MAXM17503 / MAXM17504 / MAXM17505, integrate inductors, resistors, capacitors, and high-efficiency DC-DC step-down buck regulators. They operate over a 4.5V to 60V range. Customers can start with modules and migrate to ICs for volume production.

Industrial Interface

Symmetric key-based secure authenticator provides the most secure key storage possible. The MAX66242 DeepCover® SHA-256 secure authenticator configures and collects data from any embedded system through its NFC/RFID ISO/IEC 15693 and/or I²C (master/slave) interfaces.
Contactless communication secures sensitive data with the power of SHA-256 authentication. The MAX66300 is the industry's first HF RFID transceiver with integrated SHA-256 engine for secure challenge-and-response authentication.
Highly configurable IO-Link® transceiver ensures robust communications with IO-Link sensors and actuators. The MAX14826 IO-Link transceiver supports all specified IO-Link data rates, integrates multiple protection solutions, and is ideal for Industry 4.0 applications.

Signal Chain

Programmable analog offers more versatility for industrial control and automation, IoT, base-station RF controllers, and power-supply monitoring applications. The award-winning MAX11300 mixed-signal PIXI™ input/output (I/O) brings programmability to high-integration analog applications. It is the industry's first configurable, 20-channel, -10V to +10V high-voltage mixed-signal IC.
High 20-bit accuracy ensures confidence in measurement results. The MAX11905 is the fastest 20-bit, 1.6Msps successive approximation register (SAR) analog-to-digital converter (ADC). It saves up to 91% power and up to 50% space, with the best THD of -123dB at 10kHz.
Low-noise, low-distortion drivers optimize the high speed, high accuracy of SAR ADCs. The MAX44205/MAX44206 low-noise and low-power op amps drive high-speed SAR ADCs. Their wide supply range and wide bandwidth are ideal for low-power, high-performance data acquisition systems (DAS).
High resolution ADC is ideal for instrumentation applications that require ultra-low noise. The MAX11270 is an easy-to-use, 24-bit, 10mW, 64ksps delta-sigma ADC with integrated programmable gain amplifier (PGA); it offers the highest signal-to-noise ratio (SNR) and lowest power in its class.
Get high-resolution, 24-bit accuracy for weighing mass in a wide range of industrial applications. The MAXREFDES75#, a 24-bit weigh scale reference design, features the 24-bit MAX11270 delta-sigma ADC. The MAX11270 is ideal for process control, automatic test equipment (ATE), medical instrumentation, and battery-powered devices.
Ultra-robust, 5KVrms, 4-channel digital isolators transfer digital signals between circuits with different power and ground domains, where noise isolation, ground loop mitigation, and/or safety are of concern. Industry’s only 1.8V supply capable isolators, the MAX14930 /MAX14931/ MAX14932 / MAX14934 / MAX14935 / MAX14936 deliver best-in-class propagation delay of up to 30% better and pulse width distortion of up to 50% better than competitors.

Maxim Integrated
Read more

Method for Low Power Physical Design in EDI

Today’s market increasingly expects MCU with higher performance and lower power consumption. Low power implementation becomes the key challenge in SoC. Physical design is one of the critical phase to implement it. In this paper, we provide physical implementation methods base on Cadence Encounter Foundation Flow environment. The methods focus on SOG utilization increase or gate count decrease hence to reduce power dissipation. These methods include: floorplan adjustment, GigaOpt application, leakage power optimization, new version EDI application (13.12.000) and etc. These methods have general applicability. This paper also provides some examples with power reduction results for a mixed signal design on 90nm process.

Keywords: Low Power, Cadence EDI, GigaOpt, Leakage Power

Savindar Fu, Freescale Semiconductor, Suzhou
Ken Du, Freescale Semiconductor, Suzhou
Alex Cai, Freescale Semiconductor, Suzhou

1. Freescale Kinetis Family Introduction

Figure 1: Initial Floorplan

Figure 2: Final Floorplan
Freescale Kinetis is the first broad-market mixed-signal MCU family base on the ARM Cortex™ core and the most scalable portfolio of low power, mixed-signal ARM Cortex™ MCUs in the industry.
Kinetis MCUs are built on Freescale’s innovative 90nm Thin Film Storage (TFS) flash technology with unique FlexMemory. Kinetis MCU families combine the latest low power innovations and high performance, high precision mixed-signal capability. Kinetis portfolio consists of multiple hardware and software compatible MCU families with exceptional low-power performance, memory scalability including on-chip FlexMemory/EEPROM, and peripheral integration. Families range from entry-level to highly integrated and include a wide selection of analog, human-machine interface, connectivity, and safety and security functions. Kinetis MCUs are supported by a market-leading enablement bundle from Freescale and numerous ARM 3rd party ecosystem partners.

2. Kinetis Low Power Physical Implementation

2.1 Introduction
Kinetis MCU families combine the latest low power innovations and high performance with smaller chip size. On Feb. 25, 2014, Freescale announces the KL03 MCU, the world’s smallest and most energy efficient 32-bit MCU based on ARM® technology. The increasingly low power and small chip size requirement brings big challenge for backend physical implementation.
This paper provides physical implementation methods base on Cadence Encounter Foundation Flow environment. The methods focus on SOG utilization increase or gate count decrease hence to reduce power dissipation. These methods include: floorplan adjustment, GigaOpt application, leakage power optimization, new version EDI application (13.12.000).

2.2 Floorplan Adjustment
Hard block placement is one key factor which will impact final chip utilization. It’s time worth to adjust hard placement to get a good SOG shape for easier routing and higher density.
During back-end implementation of one Kinetis project, there is routing issue with below floorplan around red area (Flash hard block) and timing requirement cannot meet with this floorplan in figure 1. It will lead to die size increase if no smart adjustment.

• Each Flash have 64bit output for read data. There are 4 Flash in chip. From floorplan view, there is no enough space for routing. This cause congestion issue.
• Flash also have high frequency requirement according to high performance definition.
There are critical paths around Flash. Long distance routing cause worse timing.

After deep analysis and much discussion, floorplan change to shape in figure 2.
• Keep enough space for Flash output routing.
• Keep enough space for Flash soft wrapper placement.
This floorplan shape makes routing and timing closure easier.

2.3 GigaOpt Application
Cadence® Encounter® Digital Implementation (EDI) system provides the most effective methodology to maximize performance, and minimize power and area for high-performance, 100M+ instance, and power-efficient designs. EDI System delivers the most comprehensive and deterministic solution for physical implementation and design closure of today’s most demanding chip designs.
GigaOpt is an ultra-fast and multi¬threaded/highly scalable optimization technology that provides better quality of results (QoR) with faster runtime. GigaOpt provides significant improvement in worst negative slack (WNS), total negative slack (TNS), and density while simultaneously reducing dynamic and leakage power across the board.

Below are different data that compares implementation with/without GigaOpt in two Kinetis projects. Tables compare the result for gate count, utilization, timing WNS/TNS and total power.
Project 1 is a small project showing in table 1. During analysis, it is split into two

Table 1: Project 1 Implementation Result
charts. One includes gate count, utilization and total power. Another includes timing information.

• Chart 1 compares three items.
- Gate count got 0.28% improvement with GigaOpt.
- Utilization got 1.3% improvement with GigaOpt.
- Total power got 14.97% improvement with GigaOpt.

Chart 1: Project 1 Items Group 1

Chart 2: Project 1 Items Group 2

• Chart 2 compares two items.
- Timing WNS got 27.17% improvement with GigaOpt.
- Timing TNS got 74.49% improvement with GigaOpt.
Project 2 is a bigger projects showing in table 2. It also is split into two charts for analysis.

Table 2: Project 2 Implementation Result

• Chart 3 compares three items.
- Gate count got 0.32% improvement with GigaOpt.
- Utilization got 1.5% improvement with GigaOpt.
- Total power got 9.47% improvement with GigaOpt.

Chart 3: Project 2 Items Group 1

Chart 4: Project 2 Items Group 2

• Chart 4 compares two items.
- Timing WNS got 20.14% improvement with GigaOpt.
- Timing TNS got 76.86% improvement with GigaOpt.

From two projects analysis, GigaOpt will help backend to improve EDI performance.

2.4 Leakage Power Optimization

Figure 3: EDI Leakage Power Optimization
For 90nm and below technologies, leakage is the main factor which dominates over the dynamic power. Use optLeakagePower in EDI to optimize total leakage power of the design by swapping gates for gates with lower leakage power without degrading timing. Use this command after the design meets timing requirements.
This command resizes only those cells that have positive slack, unless a negative target slack is specified by the user.
Based on project 1, there is different result to show enable/disable leakage power optimization during EDI implementation in figure 3.

2.5 New Version EDI Application (13.12.000)
In EDI 13 release, GigaOpt technology is the default engine for optimization, including setup/hold/power optimization.
GigaOpt simplifies the timing closure flow.
For project 1, there is comparison between EDI 11.13 and EDI 13.12 for same design and database. The result is shown in table 3.
The result analysis also is split into two groups.

Chart 5: Project 1 EDI Version Compare

Table 3: Project 1 EDI Version Compare

• Chart 5 compare three items:
- Gate count got 0.27% improvement with EDI 13.12.
- Utilization got 0.74% improvement with EDI 13.12.
- Total power got 28.02% improvement with EDI 13.12.

• Group 2 compares two items:
- Timing WNS got 0.256ns improvement with EDI 13.12.
- Timing TNS got 1.943ns improvement with EDI 13.12.

3. Issues and Workarounds
Cadence EDI System delivers the most comprehensive and deterministic solution for physical implementation and design closure. But EDI is not a impeccable tool. During Kinetis implementation there are issues which we have to find workarounds to meet the tape-out schedule.

3.1 Fix Hold Time
3.1.1 Issue Description

For timing optimization purpose, there are special delay cell which cell day is much bigger than buffer in our library. During hold time fix phase, EDI cannot smartly select delay cell firstly for big violations but inserts small buffer chain instead.

3.1.2 Workaround
During implementation, first step only use delay cells to fix big hold violations, then use buffers to fix remaining violations. This will reduce total gate count and area significantly.

3.2 Fix Transition Time
3.2.1 Issue Description

For some hard block pins with strict transition time requirement in timing model, EDI tool inserts buffer in each optimization stage, result in a redundant long buffer chain before those pins in the end.

3.2.2 Workaround
After placement, manually insert a big buffer for those pins and set don’t touch for them to prevent tool from optimizing. Check final database to find redundant long buffer chains and delete them to save gate count and area.

3.3 Critical Path Modules Placement
3.3.1 Issue Description
A part of Kinetis project have high frequency requirement. For some critical path related soft modules, EDI cannot smartly place them in a proper location thus lead to extra effort for optimization and gate count increasing.

3.3.2 Workaround
We need to provide some guidelines for tool during placement stage for better timing (createGuide, createDensityArea, createInstGroup and etc).

4. Conclusion
Freescale Kinetis projects face to embedded MCU market and there are many requirements for low power application. We use above methods in those projects with Cadence EDI new feature and new version to improve performance and also speed up run time.
Working with Cadence EDI, we will continue to improve the low power implementation flow to reduce power dissipation while keeping high performance ■

[1] Freescale,
[2] Cadence, EDI System Text Command Reference
[3] Cadence, EDI System User Guide
Read more

Missing links

How to perform a simple link budget analysis to evaluate wireless transmission using sub-GHz modules in indoor and outdoor environments.

By Pradeep Shamanna, Microchip Technology

Short-range wireless is growing in popularity in home, building and industrial applications, notably in the sub-GHz (less than 1GHz) band. This means system designers need to understand the methods, estimation, cost and trade-offs involved. Apart from the range estimation formula, it is good to understand the wireless channel and propagation environment involved with sub-GHz.
Generally, RF and wireless engineers perform a link budget when starting an RF design. The link budget considers range, transmit power, receiver sensitivity, antenna gains, frequency, reliability, propagation medium (which includes the principles of physics linked to reflection, diffraction and scattering of electromagnetic waves), and environmental factors to calculate the performance of a sub-GHz RF radio link.
Sub-GHz wireless networks can be cost-effective in any low data rate system, from simple point-to-point connections to much larger mesh networks, where long range, robust radio links and extended battery life are priorities. Higher regulatory output power, reduced absorption, less spectral pollution and narrow band operation increase the transmission range. Improved signal propagation, good circuit design and lower memory usage can reduce the power consumption, thus increasing battery life.
Usually, sub-GHz channels are part of unlicensed Industrial Scientific Medical (ISM) frequency bands. Sub-GHz nodes generally target low-cost systems, with each node costing about 30 to 40% less than advanced wireless systems and they use less stack memory. Many protocols such as IEEE 802.15.4 based ZigBee (currently, the only protocol offering both 2.4GHz and sub-GHz versions in the 868 and 900MHz bands), automation protocols, cordless phones, wireless Modbus, remote keyless entry (RKE), tyre pressure monitoring systems and lots of proprietary protocols (including MiWi), occupy this band. However, operation in the sub-GHz ISM band induces the radios to interfere with other protocols using the same spectrum, which includes a threat from mobile phones, licensed cordless phones and so on.

Link budget

Link budget is the accounting of all gains and losses from the transmitter (TX) through the medium (free space) to the receiver (RX) in a wireless system. Link budget considers the parameters that decide the signal strength reaching the receiver. Factors such as antenna gain levels, radio TX power levels and receiver sensitivity figures must be determined to analyse and estimate the link budget.
Antenna types and sizes should also be considered as well as secondary factors such as required range, available bandwidth, data rates, protocols, interference and interoperability. Receiver sensitivity is not part of the link budget but the threshold is needed to decide the received signal capability.
The simple link budget equation is that the received power (dBm) is equal to the sum of the transmitted power (dBm) and gains (dB) minus the losses (dB). By assessing the link budget, it is possible to design the system to meet its requirements and functionality within the desired cost. Some losses may vary with time. For example, periods of increased bit error rate (BER) for digital systems or degraded signal to noise ratio (SNR) for analogue systems.

Figure 1 Microcontroller to MRF89XA module interface; wireless and RF node diagram

Figure 2 Microcontroller to MRF49XA transceiver and PICtail card interface; wireless and RF node diagram

Test requirements

The Microchip MRF89XA modules and MRF49XA sub-GHz transceiver based PICtail boards can be used for the performance measurements. The MRF89XA modules are FCC, ETSI and IC certified. They differ from other embedded sub-GHz modules by providing various regulatory and modularly certified PCB antenna (Serpentine type) features. The PICtail boards are based on wire type (1/4) antenna for different frequencies, usually mounted on the development boards or daughter cards.
The hardware interface of the transceiver modules with any of the PIC microcontrollers, generally known as wireless nodes, is illustrated in Figs. 1 and 2. The wireless nodes can be realised with a combination of the PIC MCU development board and PICtail daughter board.
The range and performance experiments require at least two wireless nodes for testing. The measurement setup is done using any of the two development boards with, for simplicity, identical sub-GHz modules on each. Otherwise, a combination of these modules can be used for measurements and analysis, based on the application.

Measurement environment

Operating terrains highly impact the wave propagation. Range tests should be conducted in various indoor and outdoor environments to provide a basic understanding of the performance of the modules. The chosen environments included line of sight (LoS) on level and uneven terrain, and obstructed paths on level and uneven terrain.
The measurements are also based on PCB antenna orientation (vertical or horizontal), output power of the sub-GHz modules (maximum or default), power amplifier or low noise amplifier (enabled or disabled value), type of antenna (PCB, wire or standard) dipole, and antenna (Serpentine, wire, or whip and dipole).
The factors affecting indoor measurements include office equipment and whether there are any signals from Wi-Fi, Bluetooth or microwave in the vicinity. Concrete structures, walls, nearby glass, wood and metal can all have an effect.

Figure 3: Vertical mounting of PICtail boards
For range tests, the main differentiating factors are the module mounting, antenna orientation and the constant battery power source.
Fig. 3 shows the vertical mounting of the antenna on the baseboard. The vertical is with elevation lobe and plane; horizontal mounting is with azimuth lobe and plane.
The antenna is mounted either vertically or horizontally based on the effective output power achieved, application space requirements and constraints, such as having a strong primary lobe based on the centre fundamental frequency and secondary lobes based on its third harmonic frequency. As radio frequencies are reduced, the antenna sizes increase proportionally. The wire length in centimetres is equal to 7500 divided by the frequency in hertz. For 433MHz, this comes to 17.3cm, and for 915MHz, it comes to 8.2cm. This equation holds good for antenna wire sizes up to a quarter of the wavelength.

Range measurement procedure

To carry out range measurements, first programme the two RF and wireless sub-GHz based transceiver nodes with MiWi P2P demonstration code. Then place one RF node on a stand on a 1.5 to 2m pole after configuring a specific operating channel. By default, the wireless node is in receiving mode.
Place a similar RF node on a second stand and set for the same working channel. Make one of the nodes stationary and the other mobile. Set up the nodes and ensure they are connected to each other. Move the mobile node and test for transmission and reception. Make the measurements every 1.5 to 3m.

Figure 4: Distance measurement method
Once the critical point is attained, measure the actual and radial distance from the TX to the RX. The critical point is where the TX and RX communications become intermittent. Return about 1.5m from the critical point and check again for reliable communications. The distance measurement method is illustrated in Fig. 4 showing that the increase in range value is a function of variables with the TX module height being the most sensitive.
The packet error rate (PER) test analyses the indoor and outdoor valid data coverage between two wireless nodes. The PER test setup is similar to the open field test setup.
The PER test between two devices is done in a single iteration with a predetermined number of data packets. The ISM (IEEE 802.15.4) specification defines a reliable link as having PER below or equal to 1% for the 1000 data packets transmitted and received. PER measures the capability of a device to receive a signal without degradation due to undesirable signals at other frequencies. The desired signal’s degradation of its PER must be less than 1% or the BER must be less than 0.1%. The PER test is conducted by adding the delay between data packets, if required.
The BER measurement is done by sending the data through the wireless nodes and comparing the output to the input. Over an infinitely long period of time, the general assumption is that the data transmission is a random process. Therefore a pseudo-random data sequence is used for the BER test. It is “pseudo” random because a truly random signal cannot be created using deterministic (mathematical) methods, but few approximations of random behaviour are available to perform accurate BER measurements. The modulation modes provide good BER performance at low SNRs. However, no simple test methods exist that enable direct BER measurements.
An accepted simple method is to calculate BER from PER. The setup for measurement of the PER and BER is similar to the range measurement.
The sensitivity test setup is used to get an indication of the sensitivity limit. The input power level to the receiver is lowered through attenuators until the PER is less than 1%, and is no longer measured at the receiver. The test setup consists of two sub-GHz modules, see Fig. 5.

Figure 5: Test set up for sensitivity

The transmitting sub-GHz module is connected through an electronic attenuator to the receiving module. Both modules are connected to a PC with a USB cable or via an RS232 serial port. The PC executes the test tool with PER test scripts using the driver utility software. All the PER tests are performed without retransmission. The PER test for sensitivity provides the user with the freedom to increase the distance between the two nodes and check how far the communications can keep PER below 1% with the compensations across channels.
Sub-GHz radios can create relatively simple wireless products that can operate uninterrupted on battery power alone for up to 20 years. Sub-GHz wireless networks can be cost effective in any low-data-rate system, where long range, robust radio links and extended battery life are priorities. Higher regulatory output power, reduced absorption, less spectral pollution and narrowband operation increase transmission range. Better circuit efficiency, improved signal propagation and a smaller memory footprint can result in years of battery-powered operation.
The narrowband operation of a sub-GHz radio can ensure transmission ranges as long as a kilometre or more. This allows sub-GHz nodes to communicate directly with a distant hub without hopping from node to node. The primary reasons for the sub-GHz range performance are lower attenuation rates, lesser signal weakening, and effects such as diffraction helping sub-GHz signals bend further around an obstacle, reducing the blocking effect.
It is good to use sub-GHz ISM bands for proprietary low duty cycle links and they are not as likely to interfere with each other. The less noisy spectrum indicates easier transmissions and fewer retries, which is more efficient and saves the battery power.
Both power efficiency and system range are functions of the receiver sensitivity plus the transmission frequency. The sensitivity is inversely proportional to channel bandwidth, so a narrower bandwidth creates higher receiver sensitivity and allows efficient operation at lower transmission rates. For example, at 433MHz, if the transmitter and receiver crystal errors are both 10ppm, the error is 4.33kHz for each. For the application to transmit and receive efficiently, the minimum channel bandwidth is twice the error rate, or 8kHz, whichever is ideal for narrowband applications.

For urban environments, the use of 12dB is a good rule of thumb for predicting the required increase in the link budget to double the transmission distance. Receiver sensitivity is the first variable in a system that must be optimised to increase the transmission distance. Other variables in a system also affect distance but must be changed by a greater percentage to equal the effects offered by changing the receiver sensitivity.
Fading due to multi-path can result in a signal reduction of more than 30 to 40dB, and it is highly recommended that adequate link margin is factored into the link budget to overcome this loss while designing a wireless system ■
Read more

Wafer Level Chip Scale Package (WLCSP)

This document provides guidelines to use the Wafer Level Chip Scale Package (WLCSP) to ensure consistent Printed Circuit Board (PCB) assembly necessary to achieve high yield and reliability. However, variances in manufacturing equipment, processes, and circuit board design for a specific application may lead to a combination where other process parameters yield a superior performance. Guidelines for package performance information such as Moisture Sensitivity Level (MSL) rating, board level reliability, and thermal resistance data are included as reference.
This document contains generic information that encompasses Wafer Level Chip Scale Packages (WLCSP). It should be noted that device specific information is contained in Datasheet. This document serves only as a guideline to help develop a user specific solution. Actual experience and development efforts are still required to optimize the process per individual device requirements and practices.

Wafer Level Chip Scale Package (WLCSP)

Package Description
Wafer Level Chip Scale Package refers to the technology of packaging an integrated circuit at the wafer level, instead of the traditional process of assembling

Figure 1: WLCSP Packages Available from Freescale
individual units in packages after dicing them from a wafer. This process is an extension of the wafer Fab processes, where the device interconnects and protection are accomplished using the traditional fab processes and tools. In the final form, the device is a die with an array pattern of bumps or solder balls attached at an I/O pitch that is compatible with traditional circuit board assembly processes. WLCSP is a true chip-scale packaging (CSP) technology, since the resulting package is of the same size of the die (Figure 1). WLCSP technology differs from other ball-grid array (BGA) and laminate-based CSPs in that no bond wires or interposer connections are required. The key advantages of the WLCSP is the die to PCB inductance is minimized, reduced package size, and enhanced thermal conduction characteristics.

Typical WLCSP Configurations and Dimensions

Table 1: Die size for WLCSP Arrays at 0.40mm Pitch
Available WLCSP packages from Freescale range from 2.0 × 2.0 mm to 5.29 × 5.28 mm in size, with a standard pitch of 0.40mm and a standard solder ball diameter of 0.250mm. The physical outlines of WLCSP packages are dynamic since those depend on actual die size. Therefore, users of devices in these packages must exercise greater care in utilization than with more standardized packages. Refer to Table 1 for details regarding die sizes for standard solder ball arrays at 0.40mm pitch, which complies to JEDEC Publication 95, Design Guide 4.18 [1] and JEDEC Standard MO-211. [2]
The PCB layout and stencil designs are critical to ensure sufficient solder coverage between the package and the Printed Circuit Board (PCB). When designing the PCB layout, refer to the Freescale case outline drawing to obtain the package dimensions and tolerances.

WLCSP Construction

Figure 2: Typical Polymer-RDL WLCSP Construction
Refer to Figure 2 for a representation of a typical WLCSP package with a RDL layer between two dielectric layers. A WLCSP die has a first layer of dielectric, a Copper metal redistribution layer (RDL) to re-route the signal path from the die peripheral to a solder ball pad, and a second dielectric layer to cover the RDL metal, which in turn is patterned into the solder ball array. The solder ball is a lead-free alloy.

Process Flow
A typical WLCSP process flow is illustrated Figure 3. The illustration displays the process for a two-layer RDL process, with the RDL metal layer between two dielectric layers.

Figure 3: Typical WLCSP Process Flow

Printed Circuit Board (PCB) Level Guidelines

PCB Design Guideline

Table 2: Recommended PCB Pad and Stencil Parameters
PCB design requirements are based on IPC-A-600 [3] standards. For optimum electrical performance and highly reliable solder joints, Freescale recommends the PCB and stencil design guidelines listed in Table 2.

PCB Land Design Guidelines
Solder Mask Defined (SMD) pads are defined by the solder mask opening on the board pad as shown in Figure 4. The opening of the solder mask is smaller than the

Figure 4: NSMD and SMD Designs for WLCSP PCB Terminal
underlying copper area for soldering to the associated bump. A Non-Solder Mask Defined (NSMD) pad has a solder mask opening larger than the copper pad. There are many factors influencing whether the PCB designer uses SMD or NSMD pads. Either type can successfully be used with WLCSP packages. Freescale recommends using NSMD pads for thermal fatigue and SMD pads for drop test performance. Fillets where the trace connects to the Cu pad are recommended, especially with NSMD pads.

Via-In-Pad Structures
The need for via-in-pad structures will generally be determined by the design. Via-in-pad designs typically result in voids and inconsistent solder joints after reflow, leading to early failures. These voids are due to trapped air in the via. If via-in-pad structures must be used, it is recommended to use filled vias. As with any PCB, the quality and experience of the vendor is very important with via-in-pad designs.

Stencil Design Guideline
Due to the fine pitch and small terminal geometry used on WLCSP, particular attention must be paid to the paste printing process. In process inspection for paste height, percent pad coverage, and registration accuracy to solderable land patterns is highly recommended.

Solder Stencil Design and Fabrication
Stencils should be laser cut stainless steel with Nickel plating or electroformed Cobalt or Chromium hardened Nickel for repeatable solder paste deposition from ultra small apertures required by small pitch packages. It is recommended to inspect the stencil openings for burrs and other quality issues prior to use. Both square and round shaped apertures have been used successfully, however square shaped aperture openings provide more consistent paste printing and transfer efficiency when compared to round openings. Corners may be rounded to prevent clogging. 1:1 aperture to pad ratio is recommended for SnAgCu alloys.
For 0.40mm pitch WLCSP devices, use aperture aspect ratio of > = 0.66, with 0.25mm × 0.25mm square openings (25 micron corner radius) for improved solder paste deposition repeatability. Aperture aspect ratio is defined as the aperture opening area divided by the aperture side wall surface area.
A 0.100mm (4-mil) thick stainless steel stencil is recommended. When these stencil design requirements conflict with other required SMT components in a mixed technology PCB assembly, a step-down stencil process may be utilized in compliance with IPC-7525 [4] design standards.

PCB Assembly

Assembly Process Flow
A typical Surface Mount Technology (SMT) process flow is depicted in Figure 5.

Figure 5: SMT Process Flow

WLCSP PCB Assembly Guideline
Screen Printing: Solder Paste Material
Use of Type 4 (25 to 36 micron solder sphere particle size) or finer solder paste is recommended and a low halide (< 100ppm halides) No-Clean rosin/resin flux system be used to eliminate post-reflow assembly cleaning operations.

Component Placement
The WLCSP package is comparatively small in size. For better accuracy, it is recommended to use automated fine-pitch placement machines with vision alignment instead of chip-shooters to place the parts. Local fiducials are required on the board to support the vision systems.
Pick and Place systems using mechanical centering are not recommended due to the high potential for mechanical damage to the WLCSP device. Ensure minimal pick-and-place force is used to avoid damage, with all vertical compression forces controlled and monitored. Z-height control methods are recommended over force control. Freescale recommends the use of low-force nozzle options and compliant tip materials to further avoid any physical damage to the WLCSP device.
Use only vacuum pencil with compliant tip material whenever manual handling is required.
All assemblers of WLCSP components are encouraged to conduct placement accuracy studies to provide factual local knowledge about compensations needed for this package type. Freescale cannot anticipate the range of placement equipment and settings possible for package placement and therefore cannot make a generic recommendation on how to compensate for WLCSP interchangeability.

Reflow Soldering
Temperature profile is the most important control in reflow soldering and it must be fine tuned to establish a robust process. The actual profile depends on several factors, including complexity or products, oven type, solder type, temperature difference across the PCB, oven and thermocouple tolerances, etc. All of Freescale's WLCSP devices are qualified at Moisture Sensitivity Level 1 at 260°C. The maximum temperature at the component body should not exceed this level. Actual reflow temperature settings need to be determined by the end-user, based on thermal loading effects and on solder paste vendor recommendations.

Rework Procedure
WLCSP components removed during PCB rework should not be reused for final assemblies. Freescale follows standard component level qualifications for packages/components and these include three solder reflows survivability. A package that has been attached to a PCB and then removed has seen two solder reflows and if the PCB is double sided, the package has seen three solder reflows. Thus the package is at or near the end of the tested and qualified range of known survivability. These removed WLCSP components should be properly disposed of so that they will not mix in with known good WLCSP components.
The rework process for WLCSP devices is similar for typical BGA and CSP packages:

• To remove the faulty component from the board, hot air should be applied from the top and bottom heaters. An air nozzle of correct size should be used to conduct the heat to the WLCSP component such that the vacuum pick up tool can properly remove the component. It is recommended to apply top and bottom heaters simultaneously for 30 seconds at 300°C and 150°C, respectively. Many assembly sites have extensive in-house knowledge on rework and their experts should be consulted for further guidance.
• Once the WLCSP component is removed, the site is cleaned and dressed to prepare for the new component placement. A de-soldering station can be used for solder dressing. It should be noted that the applied temperature should not be > 245°C, otherwise the copper pad on the PCB may peel off.
• A mini-stencil with the same stencil thickness, aperture opening and pattern as the normal stencil should be used. Apply a gel or tacky flux using a mini-metal squeegee blade. The printed pads should be inspected to ensure even and sufficient solder paste before component placement.
• A vacuum nozzle is used to pick the new package up, and accurately place it using a vision alignment placement tool. A split light system that displays images of both the WLCSP leads and the footprint on the PCB is recommended.

The replaced component is then soldered to the PCB using a temperature profile similar to the normal reflow soldering process.

Moisture Sensitivity Level Rating

The Moisture Sensitivity Level (MSL) indicates the floor life of the component and its storage conditions after the original container has been opened. The lower the

Table 3: WLCSP MSL Capability
MSL value, the less care is needed to store the components. Table 3 depicts the best case MSL for each package size. All WLCSP devices at Freescale Semiconductor are MSL1, testing in accordance with IPC/JEDEC J-STD-020D. [5]
* Note: Please refer to Freescale Semiconductor web site for specific product MSL and package information, including JEDEC MSL.

Board Level Reliability

The board level reliability is usually presented in terms of solder joint life. The solder joint results in this section utilized the board layout guidelines from Section (PCB Design Guideline).

Figure 6: Example WLCSP / PCB Daisy Chain Routing (Not to Scale)
Testing Details
Samples of WLCSP in daisy chain format were used to study the solder joint reliability. BGA pairs were routed together in the WLCSP RDL layer, with a complementary pattern designed on the test PCB to provide one electrical circuit (net) through the package when the package is attached to the test PCB, as illustrated in Figure 6.

Solder Joint Reliability (SJR) Results
Assembled PCBs can be temperature cycled at a variety of temperature ranges. The most common test condition for small devices such as these at Freescale is JEDEC Condition 'G' [6] (-40°C/+125°C), with 15 minute dwell times for a typical frequency of one cycle per hour.
Freescale has the capability of continuously monitoring the resistance through a daisy chain package and its complementary test PCB. Failure is defined as resistance through the daisy chain net of 300 Ohms or greater. Daisy chain nets are tested (time zero testing) prior to temperature cycling.

Freescale continues to work on understanding and improving the solder joint reliability of WLCSP packages. From the various experiments, the solder joint reliability performances for the different package size, lead count, die thickness, and solder material are shown in Table 4. All experiments were performed using similar size test boards.

Table 4: WLCSP Solder Joint Reliability (-40°C / +125°C)

Table 5: WLCSP Solder Joint Reliability (0 / 100 °C)
An alternate condition (0/100 °C) was also evaluated. This condition employed 10 min ramp and dwell times, providing 1.5 cycles per hour. Results for this condition are shown in Table 5.


Figure 7: Underfill Selection
Data on underfill was collected on a 5.29 × 5.28mm die size with SAC1205 0.25mm diameter solder spheres. Careful selection of underfill material is critical for enhancing BLR performance of WLCSP packages. Selecting an underfill with too high a CTE can result in worse BLR performance than no underfill. Underfilling can significantly increase the solder joint reliability of WLCSP packages. A comparison of non-underfill vs underfilled results for a 5.29 × 5.28mm die size shows a 7X improvement in cycles to 1st failure (201 vs 1421).

Mechanical Drop Test
WLCSP parts were tested per JEDEC's JESD22-B111 Drop Test Specification [7]. The drop test set-up, board layout, fixtures, and criteria are all based on the JESD22-B111. All drops are carried out in the -Z direction (package down). The peak acceleration is 1500g for 0.5 ms (half-sine pulse). The resistance at time zero and still state after the drop are recorded. Resistance data was collected in-situ throughout the dropping process, with maximum resistance data recorded during the drop. The failure criteria is 100 Ohms for 200 nano-seconds, recorded 3 times during 5 consecutive drops.

Table 6: WLCSP Drop Test Results
From the various experiments, the drop test performances for the different package sizes and lead counts are shown in Table 6.

Package Thermal Resistances

The thermal performance of WLCSP is characterized using two thermal board types and three boundary conditions:

Board Types:

1. Single Signal Layer - 1s (designed per JEDEC EIA / JESD51-3 [8].
2. Two Signal Layers, Two Internal Planes - 2s2p (designed per JEDEC EIA / JESD51-5 [9] and JEDEC EIA / JESD51-7 [10]

Thermal Resistance Boundary Conditions:
1. Junction-to Ambient (Theta-JA)
2. Junction-to-Board (Theta-JB)
3. Junction-to-Case (Theta-JC)

These thermal resistances help bound the thermal problem under distinct environments.

Junction-to Ambient (Theta-JA)
Junction-to-ambient thermal resistance (Theta-JA JEDEC EIA/JESD51-2 [11]) is a one-dimensional value that measures the conduction of heat from the junction (hottest temperature on die) to the environment near the package. The heat that is generated on the die surface reaches the immediate environment along two paths: (1) convection and radiation off the exposed surface of the package and (2) conduction into and through the test board followed by convection and radiation off the exposed board surfaces. Theta-JA is reported with two parameters, depending on the board type used: R JA and R JMA. R JA and R JMA help bound the thermal performance of the WLCSP package in a customer's application.

• R JA measures the thermal performance of the package on a low conductivity test board (single signal layer - 1s) in a natural convection environment. The 1s test board is designed per JEDEC EIA/JESD51-3 and JEDEC EIA/JESD51-5. R JA helps estimate the thermal performance of the WLCSP when it is mounted in two distinct configurations: (1) a board with no internal thermal planes (i.e., low conductivity board) or (2) when a multi-layer board is tightly populated with similar components.
• R JMA measures the thermal performance of the package on a board with two signal layers and two internal planes (2s2p). The 2s2p test board is designed per JEDEC EIA/JESD51-5 and JEDEC EIA/JESD51-7. R JMA provides the thermal performance of the package when there are no nearby components dissipating significant amounts of heat on a multi-layer board.

Junction-to-Board (Theta-JB)
Junction-to-board thermal resistance (Theta-JB or R JB per JEDEC EIA/JESD51-8 [12]) is also provided for the WLCSP. R JB measures the horizontal spreading of heat between the junction and the board. The board temperature is taken 1 mm from the package on a board trace located on the top surface of the board.

Junction-to-Case (Theta-JC)

Table 7: WLCSP Thermal Performance
Another thermal resistance that is provided is junction-to-case thermal resistance (Theta-JC or R JC). The case is defined at the exposed pad surface. R JC can be used to estimate the thermal performance of the WLCSP package when the board is adhered to a metal housing or heat sink and a complete thermal analysis is done.
Table 7 has some thermal information for certain WLCSP packages [13]. All of the data was generated using Silicon (Si) die. There is an inverse relationship between the body size of the package and the thermal resistances. Large packages have lower R JMA values. The greater the body size the more PTH vias will fit under the package ■

1. Junction temperature is a function of die size, on-chip power dissipation, package thermal resistance, mounting site (board) temperature, ambient temperature, power dissipation of other components on the board, and board thermal resistance.
2. JEDEC EIA/JESD51-2 with the single layer board horizontal. Board conforms to JEDEC EIA/JESD51-3 and JEDEC EIA/JESD51-5.
3. Per JEDEC JESD51-6 [14] with the board horizontal. Board conforms to JEDEC EIA/JESD51-5 and JEDEC EIA/JESD51-7.
4. Thermal resistance between the die and the printed circuit board per JEDEC EIA/JESD51-8. Board temperature is measured on the top surface of the board near the package.

[1] JEDEC Publication 95, Design Guide 4.18, Wafer Level Ball Grid Arrays (WLBGA), Issue. A, September, 20004.
[2] JEDEC MO211, “Die Size Ball Grid Array, Fine Pitch, Thin/Very Thin/Extremely Thin Profile”, June 2004.
[3] ANSI/IPC-A-600G, “Acceptability of Printed Boards”, July 2004.
[4] IPC-7525, “Stencil Design Guidelines”, May 2007.
[5] IPC/JEDEC J-STD-020D.1, “Moisture/Reflow Sensitivity Classification for Nonhermetic Solid State Surface Mount Devices”, May 2008.
[6] JEDEC JESD-A104C, “Temperature Cycling”, May 2005.
[7] JEDEC JESD-B111, “Board Level Drop Test Method of Components For Handheld Electronic Products”, July 2003.
[8] EIA/JESD51-3, “Low Effective Thermal Conductivity Test Board for Leaded Surface Mount Packages,” August 1996.
[9] EIA/JESD51-5, “Extension of Thermal Test Board Standards for Packages with Direct Thermal Attachment Mechanisms,” February 1999.
[10] EIA/JESD51-7, “High Effective Thermal Conductivity Test Board for Leaded Surface Mount Packages,” February 1999.
[11] EIA/JESD51-2, “Integrated Circuits Thermal Test Method Environment Conditions - Natural Convection (Still Air)”, December 1995.
[12] EIA/JESD51-8, “Integrated Circuit Thermal Test Method Environmental Conditions - Junction-to-Board”, October 1999.
[13] V. Chiriac, “Wafer Level CSP Thermal Performance Evaluation”, Freescale Semiconductor, August 2008.
[14] EIA/JESD 51-6, “Integrated Circuits Thermal Test Method Environment Conditions - Forced Convection (Moving Air),” March 1999.
Read more