In our ongoing ‘On the Fringe’ series, some of TI’s brightest minds discuss today’s biggest technological trends and solving the challenges of tomorrow.

Gallium nitride (GaN) enables new power supply applications to operate at much higher switching frequencies than previously used silicon transistors at the same voltage. This means greater efficiency than could be achieved with silicon-based solutions under the same conditions.

With the introduction of a fully integrated prototype we announced today – the LMG5200– we are making it easy for engineers to design GaN into power solutions and push the limits of conventional power density expectations. Building on decades of expertise in power testing, we have done more than a million hours of accelerated testing on GaN and have built the ecosystem to enable GaN-based power designs.

GaN will find its home in power-dense places. So, it will make power supplies smaller while maintaining or improving the efficiency. It is being designed into electronic power supplies that convert electricity between alternating and direct current forms, change voltage levels, and perform a number of functions to ensure the availability of clean electrical power. For some products, it’s about performance. It just depends on the application.

This technology can impact anything you plug into a wall – personal computer adaptors, audio/video receivers, and digital TVs. Wall adapters take up significant space and are unsightly, and they also dissipate a non-trivial amount of power in the form of heat. GaN provides a significant opportunity both to shrink these and lower electric bills.

In audio applications, performance is impacted by electrical noise unintentionally injected into the audio signal. The lower capacitance of GaN can help minimize noise by minimizing parasitic ringing and optimizing switching times to minimize distortion.

In data centers and servers, GaN reduces the losses in the power supplies used to power the cloud. In addition, the ability to shrink the power solution will free up space for more processors, memory, or storage.

This is a similar concern for our customers focusing on telecommunication power supplies for network switching devices. Already there is industry momentum to investigate new, higher voltage architectures to reduce distribution losses and leverage GaN to convert in one step to lower voltages, which was previously inefficient with similar silicon solutions. For example, in base stations, you could reduce power loss by maintaining the standard 48 volts and directly convert this voltage to the level needed by the digital circuits. Today’s common architectures reduce the power from 48 volts to 12 volts and then to the lower voltage needed by the digital circuits. Now you can use fewer converters, thus reducing power loss.

Within the next few years, GaN could reduce adapter sizes while providing greater output power. The benefit will be having a wall adapter that is convenient to carry around while supporting higher capacity batteries that can be used for longer run time and bigger/better displays.

Customers will be able to use our products for a variety of automotive, industrial and wireless charging products, and it will give them much better performance. We are also engaging with military and space customers on wide temperature and radiation applications.

We can also develop new types of converters, motor drives and systems.

The LMG5200 difference

The LMG5200 prototype consists of a high-frequency driver and two GaN FETs in a half-bridge configuration – all in an easy-to-use QFN package – and allows power designers to quickly realize the true benefits of this material.

To give GaN the market it deserves, we had to make it easier for our customers to use and optimize the performance. We knew we had to do something different. By co-packaging the GaN FETs next to a high-performance driver, we were able to provide incredible performance in one module.

We also wanted to make the GaN device more intelligent. We’ve always strived to make devices smarter to reduce the complexity of a solution allowing our customers to focus on areas where they add the most value. To this end, we made the LMG5200 to reduce the effort needed to add GaN and its benefits to a power solution.

The sky is the limit with this technology, which will help customers find creative ways to get more efficient, and force us all to think about things differently. We have the right mix of system components and industry expertise to succeed in this area, and we’re accelerating the adoption of GaN in power applications with the right packaging, performance and proof of reliability, giving it the market it deserves.

Related links:
- Advancing power supply solutions through the promise of GaN

-Learn more about the industry's first 80-V half-bridge GaN FET module

What is your most pressing power supply design challenge today? What is that one thing your boss always asks for? How can you make your power supply cheaper, or smaller? How can your design be more efficient and create less heat?

If you’re considering these questions, you’re not alone. The cost/size/performance trade-off has been around for decades. It will always be around as long as we are still designing power supplies, but it’s also what drives innovation in the power-supply arena. What innovations are occurring today?

Cost: The integration of the power supply is not new. And it still makes sense most of the time. Including more functions inside the power-supply integrated circuit (IC) reduces the number of components in the bill of materials (BOM). This reduction decreases the time spent on the pick-and-place line during assembly, eliminates the costs associated with maintaining many components in a warehouse, and eliminates the cost of these components from the BOM cost.

What is new is what is being integrated. Modern DC/DC converters for portable applications like electronic point of sale (ePOS) devices and wearables integrate, in some cases, all of the components required for a given power supply. This integration includes not just small signal components like soft-start capacitors and loop-compensation passives, but the inductor and input and output capacitors.

Size: A primary driver of integration is not only cost but solution size. Smaller components are generally cheaper. This point is especially true of power-supply output filters. The easiest way to make the power supply’s output filter smaller is to increase the switching frequency. But this increases the losses, which lowers efficiency and creates excessive heat.

Performance: If performance is judged by efficiency and the corresponding heat generated, then it can be a real issue in ultra-small, high-power systems like solid state drives (SSDs) and tablets. There simply isn’t enough space to properly dissipate the heat generated, so the only solution is to create less heat by being more efficient. Power-supply IC vendors are creating ever-improving power-save modes and are even lowering switching frequencies (under some operating conditions) to get better efficiency. Furthermore, packaging technologies are improving to combine good thermal performance with better efficiency in a smaller size.

What’s your greatest power-supply design challenge these days? The technology is rapidly changing, so it’s critical to keep up with it and always look for the latest and greatest power supply for your current design. Applications engineers around the world like myself are ready to assist you in navigating through the trade-offs and compromises in every design to get you the best solution for your customers.

Additional resources:

• WEBENCH® software tools that help speed your time to market
• PMP9754 sensor node in an Internet of Things (IoT) reference design
• PMP9751 battery run-time extension for Global System for Mobile Communications (GSM) high-power applications reference design

I must apologize. I am going to approach analog-to-digital converter (ADC) noise from an analog perspective. You may be surprised about this view because this discussion usually focuses on a digital perspective, but the analog outlook does work out quite elegantly.

Delta-sigma ADCs such as the ADS1220 (Figure 1) give you detailed information about generated data converter noise. The units of this noise description, e_n, are microvolts-rms (µVrms) and microvolts peak-to-peak (µVp-p).

Figure 1: Resistive bridge measurement using the ADS1220 precision ADC

You can think of this noise as a referred-to-input phenomenon, similar to an amplifier (Figure 2). With the amplifier, the units of noise, e_n, are nanovolts per root hertz (nV/ÖHz). Over a specified bandwidth, the e_n units are microvolts-rms (µVrms). Another unit of measure for operational amplifier (op amp) noise is microvolts peak-to-peak (µVp-p).

Figure 2: Op amp circuit with a closed-loop gain of –200 V/V

Let’s now look at the referred-to-output noise of a delta-sigma ADC and op amp. Both are closed-loop systems, where the ADC processes the signal with an internal digital filter and the amplifier has an external resistor network. In both cases, the devices send the output-noise signals to the output pin. In the case of the ADC, this would be the D_OUT pin. In the case of the amplifier, this would be the V_OUT pin (and you are looking for the total noise).

Typical specifications for the delta-sigma ADC’s output noise are effective-number-of-bits (ENOB) and noise-free-bits (NFb). Equations 1 and 2 are generic formulas for these two specifications:

ENOB = log[(2*V_REF/GAIN)/( e_{n_uVrms})] / log 2                   (1)

NFb = log[(2*V_REF/GAIN)/(e_{n_uVp-p})] / log 2                        (2)

where V_REF is the data converter’s applied voltage reference and GAIN is the data converter’s gain, per an internal programmable gain amplifier (PGA).

These specifications are very important if you plan to compare ADC to ADC, but consider the power behind the noise specifications when determining the repeatability of your sensor system.

In Figure 1, the sensitivity of the load cell is 2 m/V and the maximum capacity (FSg) is 100 kg. For this system, the repeatability in grams (REPg) or minimum measurement value is equal to Equation 3:

REPg = e_n:p-p * FSg / (AVDD * Sensitivity)                 (3)

where AVDD is the voltage across the bridge as well as the ADS1220 positive analog supply.

For this ADC, at a data rate (DR) of 20 SPS, a PGA gain of 64 and in normal mode, FSg is equal to 100,000g and ADS1220µV_PP noise is 0.35 µV (per Table 1).

REPg = FSg (1 - e_n:p-p / (AVDD * sensitivity))

REPg = 0.35 µV * 100 kg / (5 V * 2 mV/V)

REPg = 0.7 gm

Table 1: ADS1220 in µV_RMS (µV_PP), normal mode

The analog point of view is useful in the oddest places. With load cells and delta-sigma ADCs, you can stretch your imagination into the analog domain and use the specifications to your system’s advantage.

Additional resources:

• Check out the ADS1220 and OPA365 data sheets.
• Read more about ENOB, signal-to-noise and distortion (SINAD) and more in this blog post.
• Bonnie, Baker. “Does ENOB tell the whole story?” EDN, April 2008
• Find more information about delta-sigma ADCs in the Best of Baker’s Best Precision Analog Delta-Sigma ADCs ebook.

Many times in life we are met with challenges that seem to come all at once. In emergency rooms, they use terms like “triage” where degrees of urgency are assigned to individual patients to determine the necessary order of care. For example, someone attacked by a grizzly bear gets treated before the child with a sore throat. Makes perfect sense, right?

As a circuit designer, we are met with similar challenges albeit less dramatic. We are also taught as a young electrical engineering student to separate sources, currents and voltages as we analyze circuits using techniques like superposition, Kirchoff’s current law and nodal analysis. All tools we use to simplify a daunting complex circuit into something more manageable to understand.

The same logic holds for designing a motor driver circuit. As the brushes in a brushed motor make and break connections with the segments in the commutator, sparks fly. We have all seen this in our handheld power drills at home. The brushed motor used in thousands of applications is a noisy beast. As the brushes wear over time, momentary shorts between power and ground can occur and cause the supply to collapse repeatedly. These very small dips (100s of nanoseconds) can cause faults in your system and overrun your microprocessor with interrupts. This can be seen in Figure 1 below. Using a differential probe between the motor supply and boar d GND, the noise is very apparent and causes under-voltage faults to occur in the system.

Figure 1

By separating the motor supply from the logic supply, this problem is solved in devices like the DRV8835 and DRV8836, DRV8837 and DRV8838 and DRV8839.These devices are specified to operate down to 1.8V on the logic supply (VCC), which is typically connected to a regulated microprocessor power rail. But, here is the solution to your problem – the motor supply (VM) is specified to operate down to 0V!! So, all those annoying supply dropouts you see in figure 1 result in ZERO faults in the logic core.

There is one more advantage that might have been overlooked. The input logic thresholds are scaled to the logic power supply (VCC).This allows the device to work with 1.8V logic which is more common these days in battery operated devices.

Complex problems are always easier to solve when you separate the variables. It is no different in the challenging motor world, separate your logic supply from your motor supply and enjoy your free time not spent on supply filtering or interrupt service routines.

Have you used this design trick before? How do you feel it will help your design?

Need more help with your design? Ask questions and get answers from TI engineers on our Motor Driver E2E forum.

Even if you don’t need the highest accuracy from your lithium-ion (Li-ion) battery gauge, some insights into the electrochemical process inside the cells will help you understand the suitability and limits of various approaches to gauging.

Gauging for Li-ion batteries is the process of estimating the state of charge in the battery. The state of charge is the remaining capacity as a fraction of the total usable capacity in the battery. The remaining capacity of a battery is typically given in milliamp-hour (mAh) or milliwatt-hour (mWh) units.

To calculate the remaining capacity, you’ll need to know the zero point at which the battery is considered empty. This is the minimum voltage to sustain a certain current for a system. Total usable capacity in the battery is just the remaining capacity once the battery is fully charged.

Many methods exist to gauge Li-ion batteries. The simplest form of gauging is to measure the voltage and associate the voltage to a pre-determined state of charge. This method can accurately determine the capacity of a fully relaxed battery; however, in a real system, getting a relaxed-voltage measurement is often difficult. The voltage-only method ignores the battery’s internal impedance. As the load or temperature changes, the loaded-voltage measurement will match the same state of charge.

Coulomb counting is another method to measure state of charge. This method typically “learns” the battery capacity in the first couple of discharges and will initialize to some initial state of charge upon first connection. The state of charge will increase or decrease depending on the direction of current flow until it reaches the zero point, or full capacity. Some coulomb-counting gauges attempt to track impedance with approximation factors that scale the capacity. Each battery in a system will undergo different use and aging so the approximations may become inaccurate after some cycling. Both of these methods treat the Li-ion battery as if it is a simplistic electrical model and measure voltage and current without attempting to model the battery’s internal dynamics.

Figure 1: Simplified lithium ion battery schematic

In order to properly gauge a Li-ion battery, you need to look at the battery as an electrochemical device and not purely an electrical device. Li-ion batteries consist of an anode, cathode and separator. Figure 1 shows a simplified schematic of a typical lithium ion battery. The cathode is typically a strong oxidizer such as manganese dioxide or cobalt dioxide that is very capable of accepting an electron. The anode is a strong reducer such as graphite or graphene, which is very capable of donating electrons. When the battery is connected to a system load, the electrons flow from anode to cathode through the system load; the Li-ions diffuse from anode to cathode internal to the Li-ion battery through the separator and solution surrounding the cathode and anode (electrolyte).

This diffusion process is responsible for the internal impedance of a Li-ion battery and is temperature-dependent. As the temperature decreases, the diffusion slows and the internal impedance increases; conversely, as temperature increases, diffusion speeds up and internal impedance decreases. The diffusion process is not only affected by temperature, but by frequency of the load and state of charge in the battery. Figure 2 and 3 show the impedance versus depth of discharge (inverse of state of charge) and impedance vs frequency respectively.

Figure 2: Depth of discharge vs battery impedance at certain temperatures. The depth of discharge is the inverse of state of charge. 100% depth of discharge corresponds to 0% state of charge. As temperature decreases, we can see the impedance increase.

Figure 3: Impedance spectra of a lithium ion battery. Starting from the bottom left, is the high frequency impedance, we move across the impedance curve until we get to the upper right, where we have the DC resistance. There are 5 regions on the plot. Region 1 and 2 correspond to ohmic contact and SEI impedance respectively. Region 3 is the charge transfer resistance. Region 4  is the distributed resistance of active material and ionic resistance of electrolyte in pores. Region 5 is the diffusion effects. Region 6 is the DC resistance at a constant load.

The equation below shows the reaction for the anode and cathode of a LiCoO₂ battery:

Ideally, as you cycle the Li-ion battery, the equation is completely reversible. But in a real Li-ion battery, the electrodes – especially the anode – will react with the electrolyte at the interface, causing the formation of a solid electrolyte interface (SEI). The SEI layer will eventually lead to an impedance increase and loss of lithium due to growth that blocks the pores of the electrode from cycling the battery. High temperatures and high states of charge will cause this growth to be elevated.

The anode can expand and contract from the diffusion of Li-ions and cause mechanical deformations that result in contact loss between the current collector and active anode material to further increase cell impedance. This can be further worsened by low states of charge or overdischarge.

The Li-ion battery is not just a simple electrical source, but a chemical power plant that requires a more complex modeling scheme than voltage, current or temperature alone. Impedance Track™ gas gauges from Texas Instruments look at the Li-ion battery using a more chemistry-based model. They use the internal impedance in all predictions of usable capacity.

Impedance Track gas gauges also use voltage, current and temperature information to determine the loss of lithium (chemical capacity or Qmax) as well as the rise in impedance throughout the life of the battery. Advanced thermal models that factor in self-heating and surface-temperature changes during charge and discharge account for the temperature effects on impedance. When an Impedance Track gas gauge is properly configured for the system it supports, it will maintain proper accuracy across the life of the battery due to the advanced nature of the model.

Additional Resources:

• SLUA450 – Impedance Track Algorithm Basics
• BMS University – Introduction to Fuel Gauging

Power engineers have worked long and hard to perfect the art of power supply design over the past several decades. In today’s world, they are tackling a new challenge: designing digital compensators for digital power supply designs. Much of the age-old control theory and analog design processes still apply in the digital world with some added idiosyncrasies. For instance, there is an inherent sampling error introduced when the analog signal is discretized by an analog-to-digital converter (ADC). Additionally, there is a phase shift caused by the delay in processing of the control law outputs. And finally, there are obvious bandwidth limitations of the digital power supply control loop as it approaches the Nyquist frequency, which is half the sampling frequency. These small changes in the system prevent the analog theory from mapping uniformly to the digital domain, causing a point of contention for die-hard analog power supply designers trying to convert to the realm of digital power supply design.

Designing a digitally controlled power supply typically involves the following steps, which are similar to the analog control design process:

1)      Design a digital compensator based on a theoretical plant transfer function.

2)      Measure the frequency response of the loop, which in this case is the digital compensator, the power stage (also known as the plant) and the feedback.

3)      Analyze the system frequency response.

4)      Based on the measured response, modify the digital compensator to optimize the gain margin, phase margin and bandwidth of the digital control loop.

5)      Repeat steps 2-4 until the power supply system is properly tuned.

The process of designing and tuning a digital power supply control loop, as Texas Instruments has streamlined with its latest powerSUITE digital power supply software design tools, is described in the following diagram:

The Solution Adapter tool allows you to adapt existing code examples from TI digital power supply kits and configure them to run on your custom digital power supply board that uses the same topology as in the TI kit. The GUI steps you through the process of selecting the solution to adapt, selecting the relevant options for that solution, and customizing those options to adapt the software solution to your custom hardware design.

 The Software Frequency Response Analyzer (SFRA) tool enables measurement of the open loop gain of a closed loop digitally controller power converter using software. This makes measurement of the bandwidth, gain margin and phase margin of your power supply design quick, easy and unobtrusive (see this blog post for more details).

 The Compensation Designer tool allows the design of different styles of compensators to achieve the desired closed loop performance. This can be done using the measured power stage or plant data from the SFRA tool or the modeled power stage as part of the Solution Adapter Tool. The coefficients that need to be programmed on the device are generated by the Solution Adapter and can be copied into the code directly.

 Learning these digital power supply design concepts can be easily applied using the new Digital Power BoosterPack, which implements a single DC-DC buck converter power stage and an on-board adjustable load (controlled by software).

 This development platform simplifies the task of digital compensator and digital control loop design and allows the power designers to focus more on the power stage design. What will you do with this new suite of software tools and development kit? Leave a comment below!

 For more information:

• Read TI’s white paper on how to choose a digital power IC
• TI’s digital power page
• TI C2000™ real-time control MCU solutions for digital power

Today is the first day on the floor at the Applied Power Electronics Conference and Expo (APEC), and we’re surrounded by power enthusiasts of every kind, from recent college graduates to esteemed experts in the field, and from analog power gurus to digital power specialists. No matter your background, the new set of tools we’re announcing today may be useful to you if you’re considering digital power for your next application.

The first tool is a low-cost Digital Power BoosterPack—a plug-in daughter card for the C2000™ Piccolo™ TMS320F28069 LaunchPad development kit that makes it easy to control a buck converter using the real-time control architecture of TI’s C2000 microcontrollers (MCUs). Developers can leverage TI’s C2000 Piccolo F2806x MCU, which integrates the C28x real-time processing core as well as sophisticated control peripherals, including tightly coupled PWMs with high resolution and fast ADCs to enable faster switching frequency power supplies. The small-form-factor, low-power Digital Power BoosterPack operates at an input voltage of 9V, which eliminates the risk associated with developing high-voltage systems.

This new BoosterPack is supported by our new powerSUITE graphical software tools, which is available within controlSUITE™--the destination for C2000 MCU documentation and code examples. powerSUITE includes several tools to help designers easily develop and modify their digital power applications and accelerate time to market:

• The Solution Adapter: Easily modify existing code examples for custom hardware using a simple GUI instead of writing development code.
• The Compensation Designer: Create compensators of different styles, providing a method to effortlessly tune control loops.
• The Software Frequency Response Analyzer: Quickly measure the frequency response of digital power converter control loops.

Want more information? Check out these resources, or leave us a question below!

• “Demystify the digital power compensator design process” with this blog post
• Read TI’s white paper on how to choose a digital power IC
• Visit TI’s digital power page
• Learn about TI C2000 real-time control MCU solutions for digital power

While at SXSW Interactive this past weekend, autonomous vehicles proved to be a hot topic with many panels discussing the future and societal implications of autonomous vehicles. Various car companies and universities showed up to give predictions and answer questions about how a world with autonomous vehicles will operate. It’s clear we can visualize a future with autonomous vehicles within the next decade – but how will we make it happen? In my Future15 talk at SXSW 2015, I highlighted emerging hardware technologies that will enable us to lessen the occurrence of car accidents, parking shortages and traffic jams.

Figure 1. Three hardware technologies work together to sense 360° around the vehicle, with built in redundancy to maximize safety.

 Innovation in hardware and software – the building blocks for these technologies – is critical to making these ideas a reality. Researchers are looking at three key technologies for sensing: camera, radar and “light detection and ranging” (LIDAR - see figure). Many consumers have already seen back-up cameras available in newer cars, including features that alert the drivers when there is an obstruction in the camera view. Radar has been appearing in new (especially luxury) vehicles for the past few years for collision and side-panel sensing as an alternative to camera technology. Radar systems send an electromagnetic pulse and makes calculations from the return pulse to determine if an object is dangerously close. It provides an additional layer of safety and can be effective when cameras may be less so – think low-visibility conditions. The third sensing component, LIDAR has been used in prototypes for self-driving cars but current LIDAR models are costly – almost as expensive as purchasing the car itself. Because of this high cost, the technology is likely new to the average consumer.

A LIDAR system will send out a pulse of infrared light using a laser diode, detect the reflected signal with a photodetector and compare the incoming and outgoing pulses to calculate the time of flight or difference in time stamps between the two signals. Knowing that light travels at – well, the speed of light – we can determine the distance over which the pulse has traveled. By shooting off thousands of pulses in a well-designed pattern, you can map the surrounding area of a vehicle to distinguish cars, people, trees and other obstacles. Currently, cost is a critical barrier to getting LIDAR in every car on the roads today. My research in Kilby Labs involves examining ways to enhance performance while reducing the costs involved, especially as the costs of the individual components like the laser and detector are limiting.  Additionally, we are also working on shrinking the form factor of the system overall, so we can embed multiple devices into the vehicle without affecting aerodynamics, or aesthetics.

Because all three technologies have strengths and weaknesses, researchers are looking closely at how to fuse the technologies together efficiently to make the best use out of the data each sensor can give us. One way to do this is to take all the available sensor data (camera images and video as well as radar and LIDAR information about the surroundings), send it to a central processing brain and then make decisions about how to affect steering, braking and other controls. While that is one method, researchers are actually now looking at distributing the processing power to the sensor nodes and performing some of the calculations at the sensor nodes. From here, extracted information, like an object’s distance from the vehicle or how fast it’s going, is sent to the central processor. This enables the processing costs to go down by reducing the amount of data sent to the central computer. Sensor fusion not only reduces system costs and increases efficiency but also adds an additional layer of safety through checks and balances and redundancy in information.

We’re well on our way to making autonomous vehicles a reality, but there’s a lot of work to do on both the hardware and software sides to bring costs down enough to integrate them into your next new car. Watch for a lot of exciting semiconductor innovations in the years to come, changing the way we think about driving and our commutes.

Fire alarms in building automation systems typically send out light in a strobe to get people’s attention.  Usually, this alarm is followed by an audible siren or message, and the light strobe continues until the alarm is over. In order to limit the number of fire alarms for an entire building, the strobe must be very bright. A simple way to achieve this is to use a light-emitting diode (LED) with very high pulse currents.

Instead of driving multiple smaller LEDs at a lower current, a single larger LED can be driven with a much higher current to create the required light output. Such pulse currents are well-tolerated by a single LED.  With their small duty cycles – typically less than 10% – the self-heating of the LED at such a high power is kept to a minimum. This prolongs the LED’s life and the fire alarm’s circuitry around it. Plus, using a larger LED typically costs less and occupies less printed circuit board (PCB) space than many smaller LEDs.

Generating such high currents can be problematic, since most LED drivers use many separate strings for driving multiple LEDs at much lower currents. Even if these strings can be combined into one, the available LED current may still be below the 3A required to drive the larger LED sufficiently for the required light output. The 12V input voltage bus is boosted to a higher voltage to drive multiple LEDs. Now, it must be stepped down to drive a single LED with a forward voltage of less than 4V. There are much fewer options available in the market for a step-down LED driver.

The TI Design, PMP9762: Efficient step-down, high-power LED Driver with Strobe and Synchronization for Fire Alarms, addresses this challenge by implementing the TPS62130 step-down converter as an LED driver. A few small tweaks to the standard step-down circuit are used to create a 3A step-down LED driver, which provides 90% efficiency. In addition to being small, efficient and able to operate off of a semi-regulated 12V input voltage, the PMP9762 supports a synchronization signal for multiple fire alarms in one building.  A master clock source can tie multiple combinations of the PMP9762 design together to produce a strobe light at the same time for the proper operation of a fire-alarm system.

In what other applications do you need a high-power strobe?

I am a mom of three children – one boy and two girls. I am blessed that all three of them are intelligent, inquisitive and creative individuals. And if any of them find an interest in science, technology, engineering or math (STEM) and want to pursue a STEM major in college and then a related career, I want all of them to have the same opportunities to follow their dreams.

But I fear that may not be the case. I think artificial barriers still stand today between many children and jobs like being an engineer or a scientist. I think the path to STEM careers is much clearer than it was just a generation ago, but I also believe we have a long way to go.

For many years, TI has recognized two things:

1. By 2020, we will have a shortage of qualified employees to fill the 1.4 million tech jobs available in the United States.[1]
2. Less than 15 percent of electrical engineers (the type we hire the most of) are women, less than 10 percent are Hispanic and less than five percent are African-American.[2] We have to do a better job of getting these three groups of people more interested in math and science.

Sometimes when I am playing with my son or doing homework with my daughters, I imagine a world where playing with Legos and building science fair projects are the first steps toward world-changing innovations, regardless of demographic.

We created a video to explain why we believe diversity fuels innovation and why it is so important to us. It explains why we have invested more than $150 million in STEM education in the last five years, why we donate our technology to schools and why TIers have spent thousands of hours mentoring the scientists and engineers of tomorrow.

(Please visit the site to view this video)

When I watched this video, I immediately started imagining that better world, and I hope you will do the same. I encourage you to watch the two-minute video and then comment below on what your world would look like, “If everyone with passion burning inside them get the same opportunities, and all minds from diverse backgrounds share an equal voice to build and solve problems and invent and innovate."

I look forward to reading your comments about the better world we can all make together if my three children, and your children, and all children, see the boundaries in front of them finally disappear.

Part of the intelligence in today’s smart grid is being able to monitor the grid accurately. What this means in real terms is measuring several parameters in the grid, including currents and voltages. Current measurement is a critical requirement in protection, substation automation and power quality. Different types of current sensors are used to measure current. The most common sensor technologies today are the current transformer (CT) and the low resistance current shunt.

A current transformer (CT) is a transformer which converts the primary current into a smaller secondary current. The CT is the most common sensor used for measurement and can measure up to a very high current while consuming very little power.

The low resistance current shunt offers good accuracy at a lower cost, and the current measurement is simple. When performing a high-precision current measurement, one must consider the parasitic inductance of the shunt.

When multiple channels need measurement, some of the benefits of replacing a CT-based solution with a shunt-based solution include a reduction in solution size, a reduction in solution weight, mitigation of cross talk, and potential increase in product life through lower mechanical issues by replacing the CT with a shunt.

One challenge with using a shunt for current measurement is that the isolation that a CT provides is not available anymore when replacing it with a shunt. Another challenge when using shunts is the ability to get very accurate (below 1%) current measurements across a wide dynamic range.

TI’s reference design on Isolated Shunt based Current and Voltage Sensing for Protection Relays (TIDA-00080) shows how the above mentioned challenges are overcome. To do this, the reference design uses the AMC1304 which is a 16-bit delta-sigma (ΔΣ) modulator with the output separated from the input circuitry by a capacitive isolation barrier that is highly resistant to magnetic interference. This barrier is certified to provide reinforced isolation of up to 7000 V_PEAK according to the UL1577 and VDE V-0884-10 standards. The output of the modulator is passed to a C2000™ Delfino™ TMS320F28377 microcontroller (MCU).

This solution enables an accuracy of 0.25% for current ranges from 0.5A to 10A and an accuracy of 1% from 10A to 200A. To achieve an extended range of current measurement, a provision to cascade multiple current channels is provided. The reinforced isolator is capable of providing up to 10 kVof surge related isolation. Voltages from 90V to 270V can also be measured with this solution.

More information can be found at www.ti.com/tool/TIDA-00080. 

(Please visit the site to view this video)

In today’s connected world software frequent field updates are necessary  to improve accuracy, or add on benefits or even fix bugs.  If you like these updates to be invisible to you, you are going to love this idea - Instant updates that do not require your software to restart or cause any glitch in the power supply. While concept is relatively simple for on the fly updates, challenge comes in to keeping the power supply in regulation during the firmware transition. In other words a seamless transition without losing any information is a key.

TI’s UCD3138 product family uses a novel method to control a power supply. The UCD3138 controller’s architecture employs programmable digital state machine hardware for implementing fast voltage/current loop control in conjunction with an integrated general purpose microcontroller free for necessary power supply housekeeping and system communication tasks. These advanced power peripherals operate autonomously from the processor, with no need to rely on slower firmware decisions or complex time-slicing of shared resources.

Figure 1: UCD3138 device block diagram 

The UCD3138 device’s topology support has been optimized for voltage mode or current mode control, half/full bridge topologies such as phaseshifted, hard-switched or resonant LLC and single or multi-phase boost power factor correction (PFC.) You can even operate the controller in a peak current mode control configuration to maintain transformer flux balance for topologies like a phase-shifted or hard-switching full bridge converter. You can ramp an internal digital-to-analog converter (DAC) at a synchronously controlled slew rate to achieve programmable slope compensation. This eliminates sub-harmonic oscillation and improves input voltage feed-forward performance.

Key requirements to perform on the fly updates are autonomous peripherals to run the control loop independently and more than one flash bank to download latest image into a redundant bank to stage a live switch. There are 3 devices in this family that support on the fly programming. UCD3138064, UCD3138A64 and UCD3138128, each include 64K, 64K and 128K Program flash respectively.

For the sake of simplicity we will be using the UCD3138064 as example to describe on the fly operation. This digital controller offers 64 kB of program flash memory in two 32 kB banks, allowing the storage of two firmware images.  Firmware bank A can represent the space for a power supply’s current (incumbent) firmware code and bank B can represent the space for the desired updated version of the code.

Using this technique you can update firmware from either primary or secondary side.

Secondary side firmware updates:

Figure 2: Switch Firmware Live

On-the-fly upgrade for secondary side can be achieved using the 2 steps below

1. Program block 2 with new firmware from the host via PMBus/serial peripheral interface (SPI)/universal asynchronous receiver/transmitter (UART) communication.
2. Stage and implement a “live switch” of execution from block A to B

Watch this video for more information.

Primary side firmware updates:

It is also possible to do live switch on the primary side and secondary side.

Figure 3: Primary and Secondary Live Firmware update

On-the-fly upgrade for primary side can be achieved using the steps specified below

1. Download the new primary image from the host into secondary controller Block B
2. Digital power controller from the secondary notifies the primary side about the updated image
3. Primary power controller prepares for a live switch
4. The secondary controller downloads the image in the primary digital controller
5. The primary controller stage makes a switch and notifies secondary controller
6. The secondary notifies the host about a successful switch.

Application Benefits

In server applications, “on the fly updates” eliminate the need to shut down, use “hot-swap”, or to physically replace a power supply due to firmware related issues or any feature upgrade, which would interrupt system operation. This reduces downtime in server datacenters that would have otherwise been incurred from upgrading power supply firmware.

Figure 4: Digitally controlled offline AC/DC power supply based on the UCD3138064

Data scrubbing is an error correction technique that uses a background task to periodically inspect main memory or storage for errors, and then correct detected errors using redundant data in the form of different checksums or data copies. Data scrubbing reduces the likelihood that single correctable errors will accumulate, leading to reduced risks of uncorrectable errors.  The on-the-fly capability afforded by UCD3138064 device architecture permits an easy implementation of data scrubbing.  

Figure 5: Memory Scrubbing on UCD3138064

For all TI digital power products and solutions, visit: ti.com/digitalpower

[1]    Digital Power Supply Controller Enables “On The Fly” Firmware Upgrades– April 2013  

Much time is spent discussing the need for more efficient use of limited energy resources, and with good reason. Energy demand continues to grow and the number of loads is predicted to climb exponentially as Internet of Things (IoT) deployment becomes real. In 2014, the International Energy Agency (IEA) published an enlightening (and somewhat sobering) document titled “More Data, Less Power.” It’s a 170-page fact-laden discussion of global IT energy usage with recommendations for managing the predicted growth of worldwide power consumption over the next few decades. The list of contributors is impressive, including the U.S. Department of Energy, tier-one telecom manufacturers, big data, big network and everyone in-between.

Before reading the report I was not aware of the IEA, how it came to be or its evolving mission. The IEA was originally created in 1974 after the Arab oil embargo, when 24 oil-importing countries joined together in order to create and maintain reliable energy (oil) supplies. Their focus today is to promote efficient energy use and control demand growth.

Among the key messages in the report is the need to look for and exploit opportunities to improve efficiency in existing and yet-to-be-deployed systems. These include the myriad devices surrounding us that are always powered, yet spend much of their time in standby. One device that consumes “vampire” power is a set-top box. The report states that these devices are idle 80-90% of the time and use only 20% of consumed energy to perform their primary functions, with the majority of the energy used to maintain network connectivity.

Another device with similar characteristics and broad deployment is a gaming console. In 2010, gaming consoles in the U.S. used 16TWh of electricity – about 1% of all residential demand. Global gaming-console usage energy consumption is expected to reach >70TWh by 2020.

That’s a lot of gaming.

So what’s the message here? It’s to keep your eyes and mind open to making designs efficient. Think out of the box. Challenge the status quo. Many tried-and-true solutions do not get a second look, despite having the potential for efficiency improvement. Every little bit helps, and sometimes better efficiency is one of the few ways to differentiate your solution.

TI eFuses are one of those little solutions that can make a big difference when deployed across millions of devices. The typical power-input stage in a 2A adapter-powered device will dissipate a little over 1.1W. Switching to a TI eFuse not only cuts that dissipation to less than 250mW, but it can be smaller, less costly, and provide superior protection and UL2367 recognition.

Keep your eyes open.

Machines are playing an increasing role in our daily lives, from semi-autonomous cars to helping doctors diagnose illnesses to managing our energy resources or financial markets. These tasks can be challenging and machines could handle these responsibilities more efficiently and reliably than a human ever could.

Some of these responsibilities, by nature, are more critical than others. If the energy grid goes down or a machine causes a mistake in the financial markets, the consequences can be serious. But a machine error related to the brakes of a vehicle, the engine of an airplane or a drug delivery system for a medical patient could have deadly consequences.

From loss of property to loss of life, the stakes to make sure machines minimize the risks associate with failure are important. This is why companies take special care when designing, testing and deploying their end product systems. Not surprisingly, the direr the consequences of failure, the more care and effort goes into the development of these machines.

As the machines take on more function, they get more complex, and with that, so do the processors that are the “brains” of the machine. With complex processors, it becomes harder for the system designers to easily understand all of the possible failures that could occur in the processor and how best to detect and manage these failures.

 Our Hercules™ microcontroller (MCU) team works hard to help ensure our MCUs and related software provide best-in-class processor support for customers' application-related uses, which can include industrial, automotive or medical functional safety systems. The portfolio includes certified functional safety development processes and devices that are certifiedby an independent third-party assessor.¹ As a result, they can significantly reduce a customer’s product certification effort. A key strategy to achieve this certification is to have a hardware architecture that includes a high level of built-in diagnostics. For example, Hercules MCUs can detect corrupt memory and provide a defined, self-correcting response. Its lock-step architecture features two redundant processor cores that execute the software and a special hardware module that compares outputs from each core to instantaneously detect faults. Many other checks are consistently being run to maintain the device’s integrity. It’s similar to your body: when getting sick, you feel pain or your temperature goes up to notify you something may be wrong. Likewise, Hercules MCUs have internal monitors that help detect when a defined mis-operation occurs within the MCU and indicate a status of this concern to the rest of the system.

Hardware isn’t the only element of a functional safety system that matters; the other key element is software. To help ensure that Hercules MCU software meets customer expectations, we have certified our software development process to meet ISO 26262 and IEC 61508. To help customers take the Hercules software through certification, we support them with a Compliance Support Package (CSP), which includes the typical test reports and safety collateral required by the standards. Add to this specialized development tools for fault injection, tool qualification and code profiling, and you see our desire to provide a comprehensive component supplier package to help our customers reach certification of their systems more quickly. For our customers, this is what really matters.

As more industries begin to rely on automated machines, where do you see a need for functional safety in the future?

1. Independent assessors of SafeTI™ products and/or processes include Exida, TÜV NORD, TÜV SÜD and UL.

I will never forget the first time I demo’d the PMBus. About eight years ago, I was visiting with a power-supply design engineer and giving a PMBus buck-controller demo. With a stroke of a PC key, I could change the power supply soft start, switching frequency or output voltage. This left the designer literally with his mouth hanging open, because typically, changing these design parameters would have required a trip to the lab and time soldering new resistor/capacitor values, then measuring the performance.

While it was amazing then, PMBus is even more popular now. More and more companies in wired and wireless communication, enterprise server and storage, and even industrial segments use PMBus power supplies. It has become so popular because of its:

• Ease of design.
• Reduction in design time.
• Ability to monitor power supplies to screen out bad boards.
• Ability to optimize power stages when using new application-specific integrated circuit (ASICs) (measuring the real current drawn and not having to over-design the output inductor and output caps).
• Ability to collect data for use in routines to perhaps improve data-center efficiency and power utilization.

There are even more benefits:

• PMBus makes it easy to create a new power-supply design without external analog components, so you don’t have to breathe in smoke from the soldering iron while you try new resistor and capacitor values on the bench.
• PMBus enables programming, sequencing, configuration, control, voltage margining, output-voltage adjustment, and parameter and fault monitoring through a graphical user interface (GUI), which allows for quick redesigns and a more intelligent response to the parameters and faults is observed.
• PMBus ICs can eliminate external hardware monitors, supervisors and temperature sensors, as well as discrete logic for delays.

Figure 1 – Turn On/Off PMBus Programming (multiple of the Soft Start time)

• PMBus power-supply ICs include nonvolatile memory (NVM), which stores the new design values in the IC. These values then become the new “default” values once the power supply is powered down and powered back up. So in a matter of minutes, you have a new power-supply design to get you to market faster.
• PMBus ICs eliminate external-voltage margining circuits, especially if there are multiple voltage-margining levels.

Figure 2 – Voltage Margin High and Low through PMBus

• The telemetry of the power-supply output voltage and current can be used for intelligent-system power management such as in cloud applications where many servers, storage boxes, base stations and switches are part of the operating environment. You can use this information to improve the power-usage efficiency (PUE), which is a measure of how efficiently the computing equipment in a data center uses energy. There’s also thermal telemetry as well as temperature information about the hottest parts of the board, which can predict faults before they happen.
• PMBus telemetry can also help optimize your design by providing you with the right power level of new ASICs or field programmable gate arrays (FPGAs) so that you can choose the best output inductor, capacitors and power-stage components.

Figure 3 – Output Voltage, Current, and Temperature Monitoring through PMBus

If you are interested in finding out more about how to program and monitor a PMBus power converter, visit the TI booth (No. 1001) at the Applied Power Electronics Conference (APEC) today. We will be demonstrating a unique design that minimizes the X-Y area on the board by placing the inductor on top of the IC, with a gap in between.

Missing APEC this year? Then check out this video, “How to create a PMBus inductor-on-top buck converter design”

 and take a look at our new PMBus Power Solutions Guide. Also, read “Synchronous Buck Converter Delivers 12 A With Fast Load Step Response” in Power Electronics magazine for more information about TI’s new TPS53915 synchronous-buck converter and “PMBus Gives Designers New Options for Meeting Adaptive Voltage Scaling Requirements” in How2Power.

Digital galvanic isolation has an important part to play in the world. It does an amazing thing: it protects precarious products from unscheduled electrical blowback like a suit of armor might against an unsuspected ricochet on the battlefield.

In addition to this protection, it also facilitates communication between devices.  Sometimes two devices might be on different ground planes but still need to be in contact with each other.  To illustrate both of these benefits, let’s look at a 3V microcontroller (MCU) that needs to turn on a 100V motor.  

Chances are that the MCU is on a digital ground, and the motor is on an analog ground plane.  If you were to connect the control lines of the motor directly to the MCU, you’d create a ground loop and ground potential difference (GPD) that may inhibit signals from the MCU passing to the motor.  Not to mention the connection alone might fry the MCU, but if for some reason the MCU’s control signals make it through this GPD and turn on the motor; the back electromotive forces (EMF) generated from the motor turning could come back and fry the MCU altogether!   If you designed your system this way, don’t fret… it’s not too late to isolate.  Good things take time, so don’t misinterpret procrastination as laziness; all engineers know that truly it is efficiency. No matter your design it’s never too late to add isolation.

Galvanic isolation comes in multiple formats as well as varying degrees of protection. Another way to look at it would be through the eyes of Goldilocks: one that is “just right” when others seem to be “too hot” or “too cold.” Starting off at the most isolation protection, you can get is reinforced isolation.  You may have heard about TI’s reinforced digital galvanic isolators, announced during the electronica trade fair in Munich.

These devices offer bulletproof ruggedized protection which is called for in some scenarios, such as harsh industrial environments dealing with very noisy signals that can spike up very quickly.  These might be seen as “too hot” in the Goldilocks scenario.  Functional isolation on the other end is the lowest form of isolation, where protection is offered but minimal levels or “too cold,” and basic isolation is in between being not too hot and not too cold in terms of protection, but just right.

Basic galvanic isolation covers you when you need more than “a little” but less than “a lot” of protection in environments where things can get noisy but not apocalyptic. For example, in any major industrial design there is most likely a power block connected to a noisy power supply. Depending on the quality and noise-rejection capability of that power supply (perhaps even inside it), you’d probably want to employ a reinforced isolator, since it could deal with the majority of the challenging electrical interference. What is then cleaned up and passed along to be used by the system itself should be OK to use, but what happens in theory doesn’t always translate in the real world. Thus, you might need a little more isolation in the rest of your system. This protects against runaway power-supply pulses or internally generated spikes from sensors or other industrial equipment like motors, fans, servo controls and protects more delicate microcontrollers or microprocessors or even a more costly field programmable gate array (FPGA) or digital signal processor (DSP).

So don’t procrastinate; isolate!  Find a galvanic digital isolator that’s just right for you. In fact, TI has a new line of basic digital galvanic isolators. Compared to our previous generation, these capacitive-based digital isolators offer:

• A 20% higher UL 1577 isolation rating of 3.0kV_RMS.
• 50% higher DIN V VDE V 0884-10 maximum-surge voltage rating of 6kV_PEAK.
• 80% lower active power.

If all of your galvanic digital isolators have been “too hot” and “too cold,” perhaps you should check out TI’s line of “just right” basic digital isolators and keep your design comfortable and covered.

Additional resources:

• Check out our digital isolators portal page.
• Read application notes on high-voltage test methodologies or understanding electromagnetic compliance.
• Watch videos explaining more abstract concepts like common-mode transient immunity and creepage and clearance.

If you don’t see what you’re looking for, try searching for asking a question in our TI E2E™ Community Industrial Interface forum.