For athletes to perform at their highest level – to run a mile in less than five minutes or leap into the air and dunk a basketball – they need strong muscles. The heart supplies blood through the veins to the muscles to enable movements such as gripping an object with your hand or moving a finger when your brain instructs it to. But without muscles, these spectacular feats would not possible.

Similarly, factory floors use a 24-V_DC power network to distribute power from power supplies (the heart) through wires and connectors (the veins) to different power switches (the muscles), enabling the movement of relays, actuators and sensors when the programmable logic controller (the brain) tells them to do so. This is evident in large machine tools that use power switches to internally transfer 24 V of power – from board to board, or externally in digital output and remote input/output modules – to distribute power to relays, actuators and sensors. These power switches enable robots, motor drives and other large machine tools to automate a factory to perform tasks at a high speed and with precision – much like the athletes we love to watch. Figure 1 is a visual representation of a factory illustrating the use of power switches in different types of factory automation equipment.

Figure 1: Illustration of power switches in different factory automation equipment

So what type of semiconductor devices act as the muscles in a factory environment? Industrial high-side switches from TI go beyond just distributing power and drive output loads – they also provide protection and diagnostic capabilities to prevent extended factory downtimes. For example, the switches include features such as:

• Adjustable current limiting for short-circuit and overload conditions.
• Integrated current sensing for load diagnostics, thus reducing downtime.
• Open-load detection for wire breaks.
• An integrated drain-to-source voltage (V_DS) clamp for inductive load driving.

One of the most important protections is current limiting, which helps keep factory automation systems with board-to-board connections and offboard loads from experiencing significant downtime due to short-circuits, which could be caused by condensation/humidity, wire breaks or short-to-ground scenarios. TI’s high-side switches also have the option for an external resistor to adjust the current limit, offering added flexibility. This flexibility makes it possible to use the same switch across a variety of output load conditions. There is no need to buy an additional device; simply change an external resistor to get a new current-limit value. Figure 2 is a block diagram of a dual-channel high-side switch with independent adjustable current limits for each of its output channels.

Figure 2: Dual-channel high-side switch, TPS272C45

Integrated current sensing offers the ability to sense the amount of current flowing through the high-side switch, which can be invaluable when debugging a machine that’s not working properly.

The switches output a current onto an external resistor that is proportional to the current passing through the internal power metal-oxide semiconductor field-effect transistor, which can then be fed to an analog-to-digital converter (ADC) in the system for processing.

Current sensing enables users to diagnose open-load situations if a wire were to break or become unplugged in a machine. This can greatly reduce debugging time for large machines that have many outputs or board-to-board connections by enabling technicians to better pinpoint which system may have the broken or missing wire, and get the system up and running again.

Additionally, current sensing can help determine whether a system is experiencing an overcurrent event or drawing an unexpected amount of current. This can alert technicians to bring a spare load or subsystem to replace the faulty one.

Providing feedback to the system that a certain load has turned on correctly is another benefit of current sensing. There are loads, such as a solenoid actuator, which have a very distinct input current profile. A typical current profile for a solenoid actuator is shown in Figure 3 below and has three components to it; a stall current, a notch current and a holding current. This profile can be captured by the high-side switches current sensing and sent to an ADC for analysis. This analysis can help to provide a feedback mechanism to the system, indicating that the load it’s powering did indeed turn on correctly and the system is running as expected.

Figure 3: Output current profile of a solenoid actuator captured by current sense of TPS274160

TI industrial high-side switches not only provide more muscle power with low drain-to-source on-resistance R_DS(ON) solutions; they also provide flexibility and intelligence in your factory systems to enable more reuse and less downtime. TI is continuing to build out its industrial high-side switch portfolio with lower R_DS(ON) switches, more feature integration and higher channel counts.

Additional resources

• Review the following data sheets, which describe the capabilities of high-side switches for use in industrial applications:
  • “TPS272C45 45-mΩ, Dual-Channel Smart High-Side Switch with Diagnostics.”
  • “TPS27S100x 40-V, 4-A, 80-mΩ Single-Channel High-Side Switch.”
  • “TPS274160x 160-mΩ Quad-Channel Smart High-Side Switch.”
  • “TPS27SA08 36-V, 10-A, Single-Channel Smart High Side Switch.”
• Review the following application reports, which describe the use of high-side switches in factory automation and motor drive systems:
  • “Enabling SPI Communication Using TI's Smart High Side Switches.”
  • “TPS272C45 Enables Smart Power Management of Remote I/Os.”

If you’re designing with controllers for industrial equipment, you’ve likely asked yourself such questions as:

• “What voltage levels do I need?”
• “Do I care about the current level?”
• “What frequency does the controller need to run at?"
• “Will this device need to withstand high temperatures and magnetic immunity?”

Some of these parameters might be more concerning than others, and their importance varies when designing string inverters, motor drives and e-meters.

When designing string inverters, some of the major challenges with pulse-width modulation (PWM) controllers are high startup currents. Low startup currents are important in string inverters because of the high bus voltage. If there is a high startup current, the startup resistors will require more power dissipation. Using fewer resistors can result in a less-reliable power supply because the operating temperature is higher. But if you use more resistors to limit the temperature rise of the system, that of course increases the number of components in the design, and the likelihood of a component failing. String inverters need to operate reliably in severe environments with high ambient temperatures and high altitudes. So having a controller with an extended operating temperature range is important.

Motor drives also utilize an inverter stage, yet have their own unique requirements. Electric motors need to function at a certain torque and speed in order to provide the necessary amount of electricity to whatever it is powering. Most often, a motor will provide an excess of torque and speed. Mechanical controls can adjust these levels, but this can cause inefficiency and wasted energy. A motor’s speed should match the process that it is performing, and an AC drive can maximize a motor’s ability to do this. AC drives can vary the speed and frequency of the motor efficiently. Power from an electrical supply goes into the drive, and the drive regulates what is fed to the motor. The power fed into the drive runs through a rectifier, converting AC power into DC power. The DC power is then fed into capacitors to smooth out the electrical waveform, and finally, into an inverter that changes the DC power into the AC power that goes into the motor. This final step is where the PWM controller comes in. Without a PWM controller, the motor can’t adjust the frequency and speed to the necessary inputs of what it is supplying.

In recent years, e-meters have become more high-tech with power-line communication. This communication carries data on a conductor that is used simultaneously for AC electric power transmission or electric power distribution to consumers. Because of these demands, new e-meters require a higher voltage and need a controller to operate. E-meters also require protection against magnetic interference, and a bias supply with proper undervoltage lockout (UVLO) limits, in order to protect the integrity of the system. A controller with a programmable frequency can have a major impact on a designer’s ability to tune the system.

Despite the unique requirements of string inverters, motor drives and e-meters, it’s possible to mitigate your design challenges with TI’s UCC28C44 family of PWM controllers, which use bipolar complementary metal-oxide semiconductor technology to enable low power consumption. This technology offers improved efficiency, faster current sensing and faster oscillator frequency. The devices have many features and performance advantages, including high (and fixed) frequency operation up to 1 MHz, reduced startup and operating current limits, and overload protection (UVLO). They also have an extended –40°C to 105°C operating temperature. Given these advantages, you can use the UCC28C44 product family for applications including switch-mode power supplies, general-purpose DC/DC or offline isolated power converters, and board-mounted power modules.

The constant push for more component integration and greater power density, on top of demanding project schedules, can leave engineers in difficult situations when designing a system power architecture. Looking specifically at test and measurement or optical module applications, the problem statement is no longer limited to the area (x-axis and y-axis) of the design; rather, it becomes a three-dimensional jigsaw puzzle, where the height (z-axis) of a design is also a constraint.

Fortunately, you can use high-power-density buck power modules to add much-needed flexibility to your power design. A buck power module integrates a buck controller, power switches, a power-stage inductor, and other passives such as high-frequency bypass capacitors and compensation components, all in a single device. In this article, I will explain how modern power modules, in particular dual-channel modules, help address challenges related to solution footprint area and z-height through multiphase operation.

Shrink your solution size

In industrial applications such as test and measurement and avionics, designers must meet stringent solution-size constraints in all three dimensions. Modern high-power-density modules help meet these requirements by offering tighter integration, unique board placement and closer placement to loads. They further save board space and reduce bill-of-materials cost by integrating passives such as high-frequency capacitors, bootstrap components, and one or more power-stage inductors. Using a bare die within the integrated circuit makes it possible to place high-frequency capacitors closer to the power stage within the module than when using a discrete converter, especially if there are printed circuit board (PCB) component clearance constraints. These solution-size benefits apply to both single- and dual-channel power modules.

Get more flexibility on the board

In applications with height constraints, low-profile power solutions can open up new possibilities in the PCB layout. The system’s physical form factor, or system enclosure, can limit the component height on the back side of a PCB, thus reserving that part of the board for low-profile components such as capacitors and resistors. You could tuck a low-profile module underneath the overhang of a heat sink of another device, like the heat sink of a field-programmable gate array or processor itself, making use of previously restricted board area. As shown in Figure 1, the TPSM5D1806 buck power module is one example of a low-profile module. At 1.8-mm tall, the TPSM5D1806 is shorter than many 1206 or 1210 ceramic capacitors.

Figure 1: Low-profile design of the TPSM5D1806 power module with surrounding bypass capacitors

The low z-height, combined with the small x-y solution size, may open up options to place the power module closer to the loads it is powering, which enables more accurate and effective regulation by reducing longer parasitic traces. You will still need to place taller solutions farther away from the load.

Take advantage of dual-phase operation

While both single- and dual-channel modules offer the benefits of smaller solution size and the increased flexibility of placing components on the board, dual-channel power modules such as the TPSM5D1806 have a unique advantage over single-channel power modules when it comes to power density. In addition to dual-output mode (see Figure 2), designers can use a dual-channel power module and connect them together into a single-output, dual-phase configuration. Splitting up the current between the two integrated inductors enables you to use vertically shorter inductors, each with lower saturation current ratings, saving space in the z-axis. In contrast, a single-phase solution would require a taller inductor to match the current capability of a dual-phase solution.

Dual-output modules also require less input capacitance by switching the channels out of phase from each other, reducing the peak and root-mean-square input currents and further saving space in the x- and y-axis. The article, “When to Use Single vs. Dual DC/DC Buck Regulators,” explains this concept in greater detail.

Figure 2: TPSM5D1806 power-module design layout in a dual-output configuration

Conclusion

Implementing dual-channel buck power modules into a design can help you take advantage of all of the power-density benefits associated with power modules, including a small solution size with a low profile, more board space and greater flexibility for other design considerations. However, dual-channel power modules offer additional power-density advantages when used in a single-output, dual-phase configuration, helping you work more easily in tight and power-dense designs while maintaining the flexibility of choosing between single-output (dual-phase) or dual-output (one phase per output) configurations.

Additional resources

• To learn about power-module operation limits, see the application report, “Understanding Power Module Operating Limits.”
• To learn about reducing power-supply noise with power modules, check out the technical article, “3 Ways to Reduce Power-Supply Noise with Power Modules.”

Within the last decade, manufacturers of wearable technologies such as smartwatches have made significant advancements in enabling users to track their personal health and fitness in real time. It’s now possible to leverage a wealth of statistics, such as step count, heart rate, oxygen saturation, workout duration and more, and track progress toward fitness goals in a variety of ways.

As smartwatches become indispensable, battery longevity and small form factor are essential consumer considerations, encouraging device manufacturers to continuously reduce the power consumption and product footprint – and causing all sorts of headaches for power-supply designers. Adding voltage supervisors to your next wearable design can help – and here are three reasons why.

No. 1: Voltage supervisors can help improve the reliability of your design

Figure 1 is a system block diagram of a smartwatch. The processor comes with multiple voltage rails that need undervoltage monitoring, which is vital to system stability.

Figure 1: Smartwatch block diagram

A voltage lower than the required minimum can lead to an unexpected overwrite to onboard memory, a system freeze or data corruption. Designers usually implement a resistor-capacitor (RC) charging circuit to prevent these types of processor errors and delay the reset signal.

Although the implementation is simple, a RC circuit has its limitations: the reset delay is unpredictable depending on input voltage conditions such as slew rate and low operating voltage. A RC circuit also lacks undervoltage monitoring.

Modern voltage supervisors such as the TLV841 offer a better solution – not only can they monitor for undervoltage conditions, they also offer precise delay time capabilities to ensure that the voltage rails are within operating ranges, thus enhancing system reliability, dependability and functionality.

No. 2: Voltage supervisors can help improve design accuracy

Today’s microprocessors need very accurate voltage monitoring. Voltage supervisors provide an increased level of precision to address any issues resulting from the low accuracy of an integrated power good signal that usually comes from a DC/DC regulator. In addition, they offer low quiescent current (I_Q) consumption. For instance, the TLV841 consumes only 125 nA. By combining low I_Q with the ability to accurately monitor a supply voltage at 0.5%, modern voltage supervisors give you the confidence to design accurate systems without affecting battery life.

No. 3: Voltage supervisors can help you miniaturize your design

An analog watch has a diameter of 38 mm to 44 mm. With limited space and the need to design several circuits to support advanced features such as a GPS module, heart rate sensors, a step counter and touch sensors, keeping the design small can be a real challenge, especially since each of the subsystems has its own required voltage rail. Manufacturers often make compromises between product features and while keeping the device size to a minimum.

Modern voltage references are available in a wafer chip-scale package, where the size of the package equals the size of the die. The TLV841 has a tiny form factor (0.73 mm by 0.73 mm) and is available in a common pitch of 0.4 mm, with a height of 0.4 mm. Adding voltage supervisors of this size to a design is crucial to a fitness watch’s operational reliability and functionality.

Figure 2 illustrates the comparison between the size of TLV841 and a commonly sized 0603 component, such as a 0.1 μF capacitor.

Figure 2: TLV841 Evaluation board

Conclusion

As consumer demands push smartwatches to their limits in regard to physical dimensions, battery life and system performance, meeting corresponding design requirements is a priority. Modern voltage supervisors open up new application opportunities, allowing you to integrate more features into your wearable device designs while improving overall system stability and reliability, and without affecting product size.

Additional resources

• Read these technical articles:
  • “How modern voltage references help shrink small designs.”
  • “Using low-I_Q voltage supervisors to extend battery life in handheld electronics.”
• Read the white paper, “Overcoming Low-I_Q Challenges in Low-Power Applications.”

The features of a modern buck converter are simply tools to get a job done – and much like the utility belt of a certain Gotham City resident, the configuration and layout of these features were designed with space and flexibility in mind. Whether you wear a cape or not, having quick and easy access to these features means more than just solving design challenges – it means doing so without “bat”-ing an eye.

Historically, power-management solutions had independent pins for every function. Soft start, power good, switching frequency adjustment, loop compensation … every feature had a pin to go with it. This approach was (and continues to be) very beneficial. For example, the TPS54620 buck converter contains a separate power-good pin, which asserts a low signal in the event of a thermal shutdown. For products operating in harsher, more volatile conditions, having access to a pin that indicates whether your device is providing steady power is a necessity.

Even in the absence of extreme circumstances, your preferences toward power density, design footprint, ease of use, and adjustability can help you select an integrated circuit (IC) that’s more optimized for your needs. Figure 1 shows how the pinout you select addresses various design challenges.

Figure 1: Pinout style vs. design challenges

One distinct advantage of using individual pinouts is being able to finely tune each feature. When every feature has its own pin, your choice of resistor, capacitor or even short to ground allows you to set features such as the soft-start time anywhere within the device’s respective operating range. In contrast, multifunctional pinout devices and trimmed devices may have limited adjustability, or in some cases have these parameters fixed to a single value.

The features we deserve, but not the ones we need right now

In some cases, you just need a silent guardian onboard – a converter that solves your design challenges with minimal oversight on your end. This is where trims become particularly useful. Device variants such as the TPS566231P and TPS566231 offer the option of choosing between a power-good pin and an adjustable soft-start pin. As you prioritize your design requirements, you might decide that power good isn’t much of a concern; you’re perhaps far more concerned about transients at startup. In that case, you could choose the TPS566231 for its adjustable soft-start (SS) capability and not worry about monitoring the signal provided by a power-good (PG/PGOOD) pin.

By limiting your pinout, you’re saving space beyond just the footprint of the IC. You are also actively reducing your external component count. Because these different trims have fixed parameters for many of the features you’d normally adjust with added external components, you’re saving a lot of time and space. Figure 2 illustrates how the TPS566231/1P/8/8P uses fewer pins without compromising device functionality. The reduced pinout of TPS566231/1P/8/8P allows for increased ease-of-use, whereas the more elaborate pinout of the TPS54620 gives you more adjustability to better optimize your power solution.

Figure 2: TPS566231/1P/8/8P pinout vs TPS54620 pinout

It’s not the pinout but the feature set that defines you

Like a watchful protector, it’s natural for designers to want to keep an eye on things. While limiting your pinout can be a lifesaver when space is tight, you can’t deny that the adjustability of parameters takes a back seat. Thankfully, there’s a solution to meet you halfway: the multifunctional pinout.

Devices such as the TPS543620 and TPS62903 offer a MODE pin, also known as a multifunctional pin, that enables the attachment of resistor configurations to select multiple parameters and compensation loops at once. In other words, it’s possible to simultaneously adjust things such as the ramp capacitor, soft-start time and current limit by connecting a few resistors to the IC. The MODE pin also saves you from having to consider the right resistor or capacitor values needed for adjusting a single parameter within the buck converter.

A data sheet of a multifunctional pinout device usually includes a table like the one in Figure 3, showcasing how different resistor values adjust different parameters within the IC.

Figure 3: How R_MODE values affect internal component selection in a given device. R_MODE is the resistor placed at the mode pin.

No one’s ever seen a headache and great design together at the same time …

As you weigh your options when selecting a device that meets your power needs, you no longer need to associate power density or ease of use with a limited feature set. Packaging innovations over the years have enabled DC/DC buck converters to offer the features you need without compromising space on the board. And while pinouts on some devices may appear limited at first, you don’t have to be the world’s greatest detective to become aware of all of the tools at your disposal. With this knowledge, it’s only a matter of time before they build a signal to shine your name in the sky.

Additional resources

• Read the technical article, “Advantages of scalability: from Peter Parker to pin-strapping.”
• Check out the white paper, “Understanding the Trade-Offs and Technologies to Increase Power Density.”

In part 1 of this series, I reviewed power metal-oxide semiconductor field-effect transistor (MOSFET) data sheets and explained what’s in the data sheet and more importantly, what’s not, while specifically looking at the temperature dependence of some key MOSFET parameters. In part 2, I’ll focus on voltage-dependent leakage currents – the drain-to-source leakage (I_DSS) and the gate-to-source leakage (I_GSS).

Why leakage currents? There are two fundamental reasons why leakage currents are important when selecting a power MOSFET for your application. First, in electronic systems, there is a green campaign to reduce wasted power, especially when the system is operating in standby mode. And second, in battery-operated systems low leakage helps maximize both battery life for primary cells and the run time between charges for secondary cells.

MOSFET leakage currents

As shown in Figure 1, the MOSFET data sheet for the CSD15380F3 specifies two leakage currents: I_DSS and I_GSS.

Figure 1: Leakage current specifications from the CSD15380F3 data sheet

The maximum leakage is specified at one voltage: I_DSS at 80% of BV_DSS (V_GS = 0 V) and I_GSS at the absolute maximum V_GS (V_DS = 0 V). I’m often asked how these parameters vary with voltage, and the answer depends not only on the applied voltage but also on the gate electrostatic discharge (ESD) structure, as detailed in the technical article, “What type of ESD protection does your MOSFET include?” As a refresher, the three types of ESD protection used in TI MOSFETs are none (lowest leakage), single-ended (lowest leakage) and back-to-back (highest leakage).

I_GSS current

In this section, I’ll present graphs showing I_GSS variation with voltage for several TI N- and P-channel NexFET  power MOSFETs with the three types of gate ESD protection. These are typical curves for design guidance only and not a guarantee of performance. TI only guarantees leakage as specified in the MOSFET data sheet.

Figure 2 shows sweeps of I_GSS vs. V_GS for a 30-V N-channel FET (NFET) and a –20-V P-channel FET (PFET) that have no gate ESD protection. The leakage is relatively flat until V_GS gets close to its positive and negative absolute maximum limits.

Figure 2: I_GSS vs. V_GS with no ESD protection: 30-V NFET (a); and–20-V PFET (b)

Figure 3 shows I_GSS for a 20-V N-channel FET and a –20-V P-channel FET with a single-ended gate ESD protection structure. The leakage current increases exponentially when the gate ESD diode becomes forward-biased. If this is likely to occur in an application, then you must use an external gate resistor to limit the current and prevent damage to the MOSFET.

 Figure 3: I_GSS vs. V_GS with single-ended ESD protection: 20-V NFET (a); and –20-V PFET (b)

The plots in Figure 4 display I_GSS for a 60-V NFET and a –8-V PFET with a back-to-back gate ESD protection structure. These devices display a symmetric leakage characteristic around V_GS = 0 V because of the back-to-back gate ESD diodes.

  Figure 4: I_GSS vs. V_GS with back-to-back ESD protection: 60-VNFET (a); and –8-VPFET (b)

I_DSS current

The other MOSFET leakage current, I_DSS, is from drain-to-source when the FET is off. The next several graphs show I_DSS vs. V_DS for TI NFETs and PFETs with the three types of ESD protection. These are typical curves for design guidance only and not a guarantee of performance. TI only guarantees leakage as specified in the MOSFET data sheet.

Figure 5 plots I_DSS for a 30-V NFET and a –20-V PFET with no ESD protection.

Figure 5: I_DSS vs. V_DS with no ESD protection: 30-V NFET (a); and –20-VPFET (b)

Figure 6 shows I_DSS for a 20-V N-channel MOSFET and a –20-V P-channel FET, with a single-ended gate ESD protection diode.

Figure 6: I_DSS vs. V_DS with single-ended ESD protection: 20-V NFET (a); and –20-VPFET (b)

The plots in Figure 7 display I_DSS for a 12-V N-channel MOSFET and a –20-V P-channel MOSFET with the back-to-back gate ESD protection structure.

Figure 7: I_DSS vs. V_DS with back-to-back ESD protection: 12-V NFET (a); and –20-VPFET (b)

Conclusion

I hope that the typical curves of I_GSS current and I_GSS vs. V_GS, and I_DSS current and I_DSS vs. V_DS will help you understand how MOSFET leakage currents vary with voltage. TI specifies and tests the maximum leakage currents at the conditions in the Electrical Characteristics data sheet. As a reminder, always use the data-sheet limits when designing with TI FETs, and if you don’t see it in the data sheet, request it from your FET vendor.

Additional resources

• Visit the TI MOSFET support and training center.

• Check out these technical articles
  • “What type of ESD protection does your MOSFET include?”
  • “What’s not in the power MOSFET data sheet, part 1: temperature dependency”

Electric vehicles (EVs) and hybrid EVs (HEVs) are changing, as are the electronics inside them. The increasing amount of electronics plays a significant role in the overall form and function of these vehicles. Drivers, however, haven’t changed; they still expect their EVs and HEVs to drive further without stopping, become more affordable, charge faster, and keep them safe. But how can designers give them more, for less?

With more stringent requirements for safety, power density and electromagnetic interference (EMI), different power architectures have emerged to address these challenges, including a distributed power architecture with an individual bias supply for each critical load.

Traditional bias power architectures in EVs

Automotive design engineers can design schemes for certain power architectures based on the EV’s power-supply requirements. The traditional approach shown in Figure 1 is a centralized power architecture, which uses one central transformer and a single bias controller to generate the bias voltages for all gate drivers.

Figure 1: A centralized architecture in an HEV/EV traction inverter

Centralized architectures have been a popular solution historically given their low cost, but this architecture can make it difficult to manage faults and regulate the voltage, along with being challenging to lay out. A centralized architecture can be susceptible to more noise as well, and have tall and heavy components in one area of the system.

Finally, as reliability and safety become a priority, a centralized architecture’s supplies lack redundancy and could result in large system failures if a single component were to fail in the bias supply. Implementing a distributed architecture for protection against power-supply failures will achieve a reliable system.

Enabling high reliability with a distributed architecture

If a small electronic component fails in a traction inverter motor while the car is going 65 mph, no one wants the vehicle to suddenly come to a complete stop or lose engine power. Redundant and backup power supplies within the powertrain have become the norm to ensure safety and reliability.

A distributed power architecture meets the reliability standards of an EV’s environment by assigning each gate driver a dedicated, local, well-regulated bias power supply in close proximity. This architecture provides redundancy and improves how the system reacts to single-point failures. For example, if one bias supply paired to a gate driver fails, the other five bias supplies remain operational, as do their paired gate drivers. If five of the six gate drivers remain operational, the motor can slow and shut down in a well-controlled manner, or potentially continue operating indefinitely. Passengers in the vehicle may not even recognize a disturbance with such a power system design.

The large height, weight and area of external transformer bias supplies such as flybacks and push-pull controllers prevent the use of a distributed architecture in lightweight electronics. The EV power system requires something more advanced – a smaller integrated transformer module such as the UCC14240-Q1 isolated DC/DC bias supply module, which integrates the transformer and components into one optimized module solution with low-height planar magnetics.

Integrating the planar transformer in an integrated circuit-sized package makes it possible to drastically reduce the size, height and weight of the power system. The UCC14240-Q1’s integration of the transformer and isolation offers easy control and low primary-to-secondary capacitance, improving common-mode-transient immunity (CMTI) in dense and fast-switching applications. Fully integrating the primary- and secondary-side control with the isolation achieves a regulated ±1.3% isolated DC/DC bias supply all in one device. By achieving 1.5 W of output power, even up to 105°C, the UCC14240-Q1 can power a gate driver in a distributed architecture, as shown in Figure 2.

Figure 2: A distributed architecture in an EV/HEV traction inverter using the UCC14240-Q1

Other considerations for driving powertrain systems in a distributed architecture

EVs require a high standard of reliability and safety, and that requirement trickles down to individual power-conversion electronics. Components must operate in a controlled and proven manner in ambient temperatures of 125°C and beyond. The isolated gate drivers are “smart” and include several safety and diagnostic features. The low-power bias supplies powering the gate drivers and other electronics in the system require advancements as well, including ways to achieve low EMI. By leveraging TI’s integrated transformer technology and a low 3.5-pF primary-to-secondary capacitance transformer, the UCC14240-Q1 can mitigate EMI caused by high-speed switching and comfortably achieve CMTI of more than 150 V/ns.

The proximity of the bias supply to the isolated gate driver in a distributed architecture ensures a simpler printed circuit board layout and better regulation of the voltage powering the gate driver, ultimately driving the gate of the power switches. These factors lead to better efficiency and reliability of the traction inverter, which can typically operate at 100 kW to 500 kW. These high-power systems demand the highest efficiency to ensure minimal heat loss, since thermal stress is one of the main culprits of component failure.

As these EV power systems move to higher power, it’s time to consider silicon carbide and gallium nitride power switches to enable smaller and more efficient power supplies. Both semiconductor technologies have several benefits, but require more tightly regulated gate-driver voltages than mature, legacy insulated gate bipolar transistors. They also require components offering low capacitance across the safety isolation barrier and high CMTI, because they switch high voltages at faster edge rates than previously thought possible.

Moving to a reliable, long-range future for EVs

Drivers will continue to demand vehicles with lower emissions, longer ranges, better safety and reliability, and generally more features for less money. Only advances in power electronics can make these demands possible in EVs, including innovations in power architectures and their associated isolated gate drivers and bias supplies.

The move to a distributed power architecture greatly increases reliability in isolated high-voltage environments, but comes with the challenge of increased size and weight from the additional components. Fully integrated power solutions such as the UCC14240-Q1 bias supply module, which switches at high frequencies, can provide both system-level space and weight savings.

Additional resources

• Check out these white papers:
  • “A High-Performance, Integrated Powertrain Solution: The Key to EV Adoption.”
  • “Power Through the Isolation Barrier: The Landscape of Isolated DC/DC Bias Power Supplies.”
• Download the UCC14240-Q1 data sheet.
• Download the UCC5870-Q1 data sheet.
• Read the technical article, “The value of a thin, isolated power solution.”

• To learn more about TI’s full portfolio of power and signal isolation technologies and how they deliver the industry’s highest isolation ratings and device lifetimes, see www.ti.com/isolation.

(Co-authored by Wenhao Wu)

Effective ways to implement backup power from supercapacitors

Many modern, smart Internet-of-things (IoT) devices that run from line power need backup power to safely power down or to perform last-gasp communication in the event of an unexpected power outage. For example, an electricity meter could share details about the time, location and duration of a power outage through a radio-frequency (RF) interface. Recently, narrowband IoT (NB-IoT) has become popular for such actions given these advantages:

• The use of existing 2G, 3G and 4G bands.
• Support from one or more operators in the Americas and European and Asian countries.
• Significantly lower power and peak currents compared to general packet radio service (GPRS).

A well-designed backup power scheme helps deliver the right amount of backup power, switches seamlessly between normal and backup operation, and can support many power outages without maintenance. In this article, we’ll present a simple method to implement a backup power scheme for NB-IoT and RF standards using TI’s TPS61094 buck/boost converter and a single supercapacitor. We’ll also compare the TPS61094-based solution with existing TI reference designs.

Watch the TPS61094 in action

See how the TPS61094 boost converter integrates a buck mode for supercapacitor charging while maintaining ultra-low IQ in our video "Achieve ultra-low 60-nA IQ with the TPS61094."

Backup power for NB-IoT

Table 1 shows the current consumption during different NB-IoT operating modes over time. Peaking at 310 mA in data transmit mode for 1.32 s, the load varies significantly with operation mode changes. The average current consumption in the entire process is 30 mA for 80 s – a load duration that requires enough backup power and seamless power switching when the main power grid is suddenly down. The TPS61094 60-nA quiescent current (I_Q) bidirectional buck/boost converter enables reliable and simple backup power designs, while serving as a one-chip solution that achieves supercapacitor charge and discharge without extra circuitry.

* Customer and end equipment specific. Can be several minutes to days.

Table 1: NB-IoT load profile example from Saft Batteries

To implement an effective backup power circuit with a single supercapacitor and the TPS61094, Figure 1 shows how we configured the TPS61094 evaluation module (EVM) to support enough backup power for the NB-IoT load profile in Table 1.

Figure 1: TPS61094 EVM backup power configuration

When system power is on, the TPS61094 enters Buck_on mode, which turns on the bypass field-effect transistor (FET), supplies the supercapacitor with 500 mA of constant current, and stops charging when the voltage across the supercapacitor is 2.5 V. V_SYS powers V_OUT directly. When a power outage causes V_SYS to drop, the TPS61094 automatically enters Boost_on mode, turning off the bypass FET and supplying V_OUT from the energy stored in the supercapacitor.

Figure 2 shows oscilloscope measurements for a complete backup power cycle. V_IN represents the system voltage from the grid. V_OUT is the output voltage from the TPS61094 and V_SUP is the supercapacitor voltage. I_OUT shows the load current consumption. In our example, the load is 100 mA, which is 3.33 times the average current consumption of the load profile. We increased the load in order to determine how the TPS61094 switches input power when the grid is down in a more extreme load condition.

When system power goes down suddenly, the TPS61094 immediately enters Boost_on mode and regulates V_OUT using power from the supercapacitor. The buck/boost converter supplies the required output current for 254.5 s, which equals 11.5 NB-IoT transactions. The TPS61094 discharges the supercapacitor until its voltage drops to 0.7 V; at that point, the device enters shutdown mode until the system V_IN returns. In Buck_on mode, the TPS61094 seamlessly charges the supercapacitor with constant current. As you can see in Figure 2, the switchover between supercapacitor discharge and charge is very smooth.

Figure 2: TPS61094 power-cycle measurements

Other backup power implementations

There are also other solutions available to you, each with advantages and drawbacks. The Supercapacitor Backup Power Supply for E-Meters Reference Design, uses discrete circuitry to charge the supercapacitor and the TPS61022 boost converter to boost the supercapacitor voltage to a higher system voltage when the grid is down. The TPS61022 output current capability is higher than it is for the TPS61094 solution, but will require more external components.

Another approach is the Supercapacitor Backup Power Supply with Current Limit and Active Cell Balancing Reference Design, which uses the TPS63802 buck-boost converter as a supercapacitor charger and voltage regulator and eliminates extra discrete charging circuitry. It still needs additional external components to achieve power ORing, charge current limit and supercapacitor terminal voltage settings, however.

Table 2 lists the most important characteristics of each backup power approach.

*For the TPS61094 and TPS61022, the minimum V_IN is 0.7 V. For the TPS63802, the V_IN is 1.3 V.

Table 2: Backup power solution overview

Conclusion

Low-power, wireless standards are becoming increasingly popular. For high-integration, simple design and the best light-load efficiency, the TPS61094 is a good choice for backup power applications with LTE-M, Lora, Bluetooth and other emerging wireless interfaces.

If you need more output current, the e-meter or current limit reference designs are very effective solutions. Although they require more discrete components, they can support higher-power RF transmissions such as GPRS.

Additional resources

• Download the TPS61094 data sheet.
• Check out the TPS63802 hardware development kit.
• See the application brief, “Solving the Backup Power Supply Challenge for Electricity Meters with High-Power Wireless Communications.”
• Read the Analog Design Journal article, “I_Q: What It Is, What It Isn’t, and How to Use It.”

(Co-authored by Alex Pakosta)

Lithium thionyl chloride (LiSOCI₂) batteries are popular in smart flow meters because they provide higher energy density and a better cost-per-wattage ratio than battery chemistries such as lithium manganese dioxide (LiMnO₂). One disadvantage of LiSOCl₂ batteries is poor response to peak loads, which can result in a decrease of the usable battery capacity. So in this article, we’ll describe an effective method to decouple peak loads from the battery, in the range of a few hundred milliamperes, that can help increase battery life.

Maximizing the usable battery capacity is important because it enables the system design to support:

• More meter readings and data transmissions from the same battery.
• A longer lifetime from the same battery.
• A smaller battery for the same operating lifetime.

The overall effect minimizes battery and maintenance costs, as well as development costs, by enabling more reuse of a single flow meter design across more kinds of flow meters

Watch the TPS61094 in action

See how the TPS61094 buck/boost converter maintains ultra-low IQ while integrating supercapacitor charging support in our video "Achieve ultra-low 60-nA IQ with the TPS61094."

The design challenge: extend the battery life

A successful meter design needs to sustain a long operational time (>15 years) while enabling functionalities such as valve control, data recording and data transmission. Extending battery life is an effective way to increase meter operational time. If you connect the battery to the load directly without any power buffer in between, however, the meter’s complex load profile may deteriorate the battery’s lifetime performance.

Based on the current level, you can divide the load consumption profile of a standard meter into standby mode, middle-stage mode and active mode. Each mode influences battery life differently:

• Standby mode consumes 5 µA to 100 µA. It is mainly quiescent current (I_Q) from metrology, microcontroller and protection circuitry. Although the absolute value is very small, it is typically the main contributor to meter lifetimes. In standby mode, the I_Q of any connected DC/DC converter should be in the nanoampere range, with the leakage of any power buffer small in order to improve efficiency.
• Middle-stage mode consumes 2 mA to 10 mA. The analog front end in RX stage usually contributes to this load. The power buffer’s efficiency is important to minimize energy loss in this mode.
• Active mode consumes the highest current. In active mode, the load usually comes from the driving valve and analog front end in TX stage, which needs 20 mA to several hundred milliamperes. Directly drawing this current from a LiSOCl₂ battery causes severe capacity derating.

Table 1 demonstrates the Saft LS33600 battery’s capacity derating vs. nominal capacity of 17 Ah at different load and temperature conditions. At an operating temperature of +20°C, a 200-mA load current leads to a 42% capacity degradation. Therefore, the battery should never directly supply the load. Only by employing a low-leakage power buffer can you limit the peak current to less than 10 mA.

Table 1: Characteristics of capacity vs. current for the LS33600 from Saft Batteries

TI’s 60-nA I_Q buck/boost converter, the TPS61094, helps extend battery life while maintaining excellent efficiency over standby, middle-stage and active modes. The TPS61094 has three main benefits:

• Ultra-high efficiency in a wide load range. The TPS61094 has >90% average efficiency for loads from 5 µA to 250 mA under conditions of V_OUT = 3.3 V and V_IN>1.5 V. This enables an efficient power supply in most flow-meter use cases.
• Limits the peak current drawn from the battery. The TPS61094 can limit its peak input current when it is working in Buck_on mode when charging the supercapacitor, and also in supplement mode when it is supplying a heavy load on V_OUT with the battery. Figure 1 illustrates the configuration of the TPS61094, while Figure 2 shows the battery’s peak current when there is a 200-mA and 2-s load pulse on V_OUT. In phase 1, where the load is heavy, the peak current is limited to 7 mA. After the load is released in phase 2, the device is charging the supercapacitor with a 10-mA constant current. When the supercapacitor voltage charges back to 2.0 V, the device stops charging but still stays in Buck_on mode.

Figure 1: Configuration of the TPS61094

Figure 2: Oscilloscope result of battery peak current at heavy load

• Unchanged available energy from the supercapacitor over the temperature range. Typically, using hybrid-layer capacitors (HLCs) or electric double-layer capacitors (EDLCs) as power buffers will improve pulse-load capability. The energy stored in these passive components depends on the battery voltage, however. When the temperature decreases, the battery voltage also goes down, which deteriorates the HLC or EDLC’s pulse-load capability and increases the battery’s supply current. The TPS61094 eliminates this issue by keeping the voltage on the supercapacitor stable, regardless of temperature.

The usable energy in the supercapacitor is defined by the capacity of the supercapacitor, the set maximum voltage across the supercapacitor and the undervoltage lockout of the TPS61094. The more usable energy that a supercapacitor has, the longer the operating time with a continuous, heavy load.

Figure 3 shows a power-buffer solution using the TPS61094 or only supercapacitors, respectively. For the TPS61094 solution, the supercapacitor voltage is set to 2 V. By supplying a continuous load, the TPS61094 can draw power from the supercapacitor until 0.6 V. Therefore, it is possible to calculate the available energy on the supercapacitor with Equation 1:

Equation 1

where ŋ is average efficiency of the converter.

In the worst case of –40°C, the TPS61094 has an average efficiency of 92% at 150 mA for an input voltage from 2 V to 0.6 V. Equation 2 shows the calculated result:

Equation 2

Figure 3: TPS61094 vs. HLC/EDLC configuration

For HLC or EDLC solutions, the available energy changes following the battery voltage. For a 10-mA current at –40°C, the LS33600 voltage reduces to 3 V. Equation 3 calculates the available energy:

Equation 3

Comparing results between Equations 2 and 3, the TPS61094 solution has double the available energy of the HLC and EDLC solutions. This means more energy can be delivered to loads, and lowers the peak current drawn from the battery under extreme conditions. For example, if there is a 200-mA load at 3.3 V to drive a valve, an HLC or EDLC solution can only support the load for 2.8 s. The TPS61094 buck/boost converter with an integrated supercapacitor can support the load for as long as 7.8 s, assuming that the power buffer supplies all of the load.

Conclusion

The complex load-consumption profile of flow meters requires a power buffer to help extend LiSOCl₂ battery life. With excellent efficiency over wide operating conditions, the TPS61094 is a good choice to remove lifetime extension challenges. By limiting the peak current drawn from the battery, this buck/boost converter maximizes its capacity and raises the supercapacitor’s available energy, enabling the system to operate longer in low-temperature conditions than an HLC or EDLC solution.

Additional resources

• Check out the TPS61094 data sheet.
• Read the application note, “The Long-lifetime, cost-competitive solution in smart meters based on TPS61094.”
• Read the Analog Design Journal article, “I_Q: What It Is, What It Isn’t, and How to Use It.”
• Read the article, “5 best practices to extend battery life in flow meters”

A common definition of quiescent current (I_Q) is the current drawn by an integrated circuit (IC) in a no-load and nonswitching but enabled condition. A broader and more useful way to think about it is that quiescent current is the input current consumed by an IC in any number of its ultra-low-power states.

For battery-powered applications, this input current comes from the battery, so it determines how long the battery operates before it either needs recharging (for rechargeable batteries, such as lithium-ion (Li-Ion) or nickel metal hydride (Ni-MH)) or replacing (for primary batteries, such as alkaline or lithium manganese dioxide (Li-MnO₂)). For battery-powered applications that spend a large amount of their time in standby or sleep mode, I_Q can impact the battery’s run time by years. For example, using an ultra-low-I_Q buck converter like the 60-nA TPS62840 to power an always-on application, such as the smart meters shown in figure 1, enables 10 years of battery run time.

Figure 1: Smart meters

I_Q also impacts the battery run time of applications that we interact with on a daily basis. Perhaps you’ve purchased a smart watch, only to find that it needed to be charged for an hour before you could use it. Or maybe you always carry a physical key for your house in case the battery in your smart lock, like the one shown in Figure 2, dies. Both of these scenarios also relate to I_Q.

Figure 2: Smart lock application

In this article, I’ll explain three of the most commonly used DC/DC converter data sheet specifications related to I_Q– shutdown current, nonswitching I_Q and switching I_Q– and how these specifications impact system power consumption.

Overcome low-Iq challenges in low-power applications

Read our white paper, “Overcoming Low-IQ Challenges in Low-Power Applications,” for tips on how to extend battery life while also providing higher performance.

Shutdown current

Shutdown current is measured when the IC is turned off or disabled. Given this, you might think nonswitching I_Q should always be zero. In reality, some ICs exhibit leakage currents in this state, while others actually have internal circuitry that consumes a small amount of current to maintain housekeeping functions even while the IC is disabled.

Think about consumer electronics sitting on a store shelf. The reason why your smart watch might not work out of the box has to do with each of its IC’s shutdown current specifications, like those shown in Figure 3. When an end product is sitting on a store shelf or on a higher shelf in a warehouse (where the temperature is likely elevated, leading to a faster battery drain), most DC/DC converters, for example, are in a shutdown state. So, even though the DC/DC converters are disabled, the battery is slowly discharging.

Figure 3: Battery discharge current in ship mode for the BQ21061

Some ICs have multiple shutdown states, such as the 2-nA ship mode of TI’s BQ25120 battery charger or the 4-nA bypass mode of the TPS61094 boost converter. In these advanced shutdown states, usually a very limited subset of the device’s functionality remains active in order to draw the minimum amount of I_Q. Compared to the 700-nA I_Q in the BQ25120’s high-impedance (shutdown) mode and the 200-nA I_Q in the TPS61094’s shutdown mode, ship mode and bypass mode extend the battery run time by 350 and 50 times, respectively.

Nonswitching I_Q

Nonswitching I_Q is when the IC is enabled, in between switching pulses, and without a load. This parameter is found in most switching DC/DC converter data sheets because it can be easily tested on production automated test equipment.

While non-switching I_Q provides an apples-to-apples comparison between different ICs, two shortcomings prevent it from being the best estimate for battery run time: the non-switching I_Q is not the same as the battery current drawn, and many ICs draw their I_Q from both the input voltage and the output voltage. However, since the output voltage and its I_Q ultimately come from the battery at the input, additional conversions or measurements are necessary in order to get an equivalent I_Q from the input source – you can’t just add the two I_Q currents to get the total battery current drawn. For example, the TPS61099 boost converter consumes a 400-nA I_Q from V_IN and a 600-nA I_Q from V_OUT, but the no-load input current consumption is about 1.3 µA and not 1 µA.

Switching I_Q

Switching I_Q, which goes by many different names: operating I_Q, standby current, sleep-mode current, no-load input current, ground current for low dropout linear regulators (LDOs) and so forth, is the real, measured input current that occurs when the IC is operating without delivering any load current. Since it is measured under real-life conditions and not on the production line, the IC occasionally switches to overcome losses and replenish leakage at the output.

It is the best estimate of the battery current drawn under no load and appears in many data sheets, such as the 60-nA switching I_Q of the TPS62840, as shown in Figure 4.

Figure 4: A 60-nA I_Q DC/DC converter

Using low-I_Q DC/DC converters is critically important to achieving the desired battery run time for applications that spend most of their time in a very-low-power state. For example, smart locks spend most of their time in a very-low-power state waiting for a phone to send the code to open the lock. If the switching I_Q is too high, most of the battery’s energy is used while waiting instead of being used to open or close the lock.

Conclusion

This article provided a brief look into how I_Q is commonly specified in data sheets and how it impacts  battery run time. For more detailed technical information about I_Q, see the white paper, “Overcoming Low-I_Q Challenges in Low-Power Applications” read my Analog Design Journal article, “I_Q: What it is, what it isn’t, and how to use it,” or take an even deeper dive into the topic by watching our low-I_Q training series.

The increasing number of electronic circuits in vehicles is placing greater demand on the amount of power consumed from the battery. The battery supply remains on even when the car is parked or turned off, in order to support such features as remote key entry and security.

Since all vehicles run on a finite battery supply, it’s important to find a way to add more functionality –particularly when designing automotive front-end power systems – without adding significant drain on the battery. The need to comply with stringent electromagnetic compatibility (EMC) standards such as the International Organization for Standardization’s ISO7637 and the LV 124 standard, which was established by German automotive manufacturers, directly affects the overall design of front-end reverse battery protection systems. Several original equipment manufacturers specify the total current consumption when the vehicle is in a parked state at <100 µA per electronic control unit (ECU) in 12-V battery-powered systems and <500 µA in 24-V battery-powered systems. 

In this article, I will review three approaches to designing a low quiescent-current (I_Q) automotive reverse battery protection system.

Using a T15 as an ignition or wakeup signal

The first approach to designing a low I_Q reverse battery protection system is to use a terminal 15 (T15) as an ignition or wakeup signal. The T15 is a wire terminal that disconnects from the battery when the vehicle’s ignition is off. Using a T15 as an external wakeup signal is one of the most traditional ways to operate ECUs in sleep or active mode. Figure 1 shows one example.

Figure 1: Reverse battery protection in an automotive ECU using a T15 for wakeup

When the ignition is on, the T15 connects to the battery voltage (V_BATT) potential, which drives the enable (EN) pin of the ideal diode logic high.   The ideal diode controller, in active mode, while enabling the charge pump, control and field-effect transistor (FET) driver circuits, actively controls the external FET for ideal diode operation. When the vehicle is being parked, the T15 drops to 0 V, and the ideal diode controller responds by entering a shutdown state, which turns the charge pump and control blocks off, resulting in an I_Q consumption of < 3 µA. In this mode of operation, the external FET turns off and the body diode of the FET forms the forward conduction path to power the load. This scheme requires an additional wiring line to the ECUs.

Using the system’s MCU and CAN wakeup signal

The second approach is to use the system’s microcontroller (MCU) and Controller Area Network (CAN) wakeup. In many cases, the communication channels of the system make a low-I_Q shutdown mode possible. Figure 2 shows an example system design using this approach.

Figure 2: Low I_Q reverse battery protection using an MCU and CAN wakeup signal for enable control

The CAN transceivers in the vehicle translate messages from the communication bus to their respective controller – typically an MCU. The transceivers can indicate when functions are unnecessary by issuing a command to go into a standby mode until they are awakened. They then relay the message instructing the controller to place the system in a low-power state – in this case, by driving the EN signal of the ideal diode controller logic low. With more advanced transceivers and system-basis chips, one device can handle multiple functions of this process, transitioning to a low-power state or waking up.

This scheme requires an internal control signal from an MCU (via CAN control).

Using an always-on ideal diode controller

A third approach is to use an always-on ideal diode controller. Imagine a system design that does not require control signals to enter low-power states. This would eliminate the need for additional wiring or system software dependencies while allowing the ideal diode controller to always be enabled, even during sleep mode. This type of system design is possible using a low-I_Q ideal diode controller such as the LM74720-Q1, LM74721-Q1 or LM74722-Q1, shown in Figure 3. These devices integrate all of the necessary control blocks for an EMC-compliant reverse battery protection design, along with a boost regulator to drive the high-side external MOSFET, which results in  27 µA of I_Q during normal operation. Learn more in the application note, "Basics of Ideal Diodes."

Figure 3: Using an always-on low-I_Q ideal diode controller without external enable control for reverse battery protection

These ideal diode controllers support reverse battery protection with active rectification and load-disconnect FET control with a back-to-back FET topology to protect the downstream during system faults such as an overvoltage event, as shown in figure 4.

Figure 4: Reverse battery protection in a 24-V automotive ECU using the LM74720-Q1

With an adjustable overvoltage protection feature, you can use 50-V-rated downstream filter capacitors (instead of 80- to 100-V-rated capacitors) and 40-V-rated DC/DC converters (instead of 65-V rated converters) for a 24-V automotive battery input-based system design.

The LM74720-Q1 and LM74721-Q1 provide fast reverse (0.45 µs) and forward comparators (1.9 µs), and a strong 30-mA boost regulator to support and enable resilient and efficient active rectification during automotive AC superimposed testing up to a 100-kHz frequency. The LM74722-Q1 offers two times faster rectification speed than the LM74720-Q1 and LM74721-Q1 devices, with a forward comparator response of 0.8 µs, enabling active rectification frequency up to 200 KHz. The LM74721-Q1 features an integrated drain-to-source voltage (VDS) clamp that enables transient voltage suppressor (TVS)-less reverse battery protection design, resulting in denser system solutions. Read more about active rectification in our application report, “Active Rectification and its Advantages in Automotive ECU Designs.” 

Conclusion

The LM74720-Q1, LM74721-Q1 and LM74722-Q1 low-I_Q always-on ideal diode controllers enable you to design an automotive reverse battery protection system without the need for external enable control. With low I_Q, back-to- back FET driving capabilities and over-voltage protection, these ideal diode controllers enable designs to use downstream components such as capacitors with lower voltage ratings, and reduce the size of printed circuit boards for space-constrained ECUs.

The world is a noisy place – and power supplies are no exception. In the pursuit of higher efficiency, power converters switching at faster and faster speeds create unintended issues, including increasing the system’s susceptibility to transients and noise.  It’s important to consider this susceptibility when choosing how to design your power supplies, and what components to design them with.

Where does the noise come from?

Where are the transients and noise coming from? Apart from the general noise found in many electrical systems, the high-frequency power supply itself often ends up generating them.  Take a look at Figure 1, which shows a controller and a gate driver driving a metal-oxide semiconductor field-effect transistor (MOSFET).

Figure 1: Transient current (di/dt) effect on the inputs of a gate driver in a power converter

As the gate driver switches the FET on and off, transients and noise are likely to occur. Parasitic trace inductances and the MOSFET’s source inductance, when combined with the fast switching of today’s power supplies, can lead to a situation known as ground bounce at the gate-driver’s integrated circuit (IC). Ground bounce is when the parasitic inductance of a system causes the ground of the IC to shift away from the ground of the system. The IC will register a 0-V signal as a negative voltage, which can cause damage or false logic outputs.

Not only is ground bounce a possible issue for the inputs, but the high switching speeds of many modern power supplies can also introduce negative voltage transients. Looking again at Figure 1, a large di/dt during switching that is proportional to the parasitic inductance (L_SS) generates a negative voltage spike (V_n) . These negative transients are proportional to both parasitic inductances and frequency, so they grow more troublesome as the switching frequency increases.

Voltage transients at the driver’s output also cause issues. Power supply output load changes are the most common cause of voltage transients. If these transients exceed the maximum voltage of the gate driver, they could end up damaging the device. Noise resulting from transients could even result in an output error, causing the driver to change its output regardless of the control signal, as shown in Figure 2.

Figure 2: Driver response to noise on the outputs

The existing solution

What can you do to address these transients, along with other noise? First of all taking the time to ensure proper board layout is important. Placing the driver as close to the switch as possible will help reduce the parasitic inductances. Doing so will not make the problem go away, however; it will just make it more manageable. In the past, a common solution to transients that could cause damage on the driver’s inputs or outputs was to add clamp diodes to hold the voltage above or below certain levels. Figure 3 shows an example of how to place these diodes.

Figure 3: Clamp diodes on a gate driver’s inputs and outputs

The diodes on the input hold the input signal above ground, while the diodes on the output hold the voltage below the supply voltage (V_DD). While effective, this is a costly solution, adding as many as six extra components to the power supply. Not only do these diodes add to the overall price of the system, they take up valuable space on the board.

Using 30-V gate drivers in your power-supply design

Newer gate drivers such as TI’s UCC27614 and UCC27624 can handle the noise and transients resulting from today’s higher-frequency power converters. These drivers offer –10-V negative voltage handling, which means that they can withstand the negative voltage shift brought on by ground bounce or negative voltage spikes caused by input transients without the need for external components. Not only do these drivers have the ability to handle lower voltages than many low-side drivers, they also have a 30-V maximum V_DD. This is an important specification, because if the gate driver has enough headroom, it can handle noise and transients on its outputs without requiring clamp diodes to hold the output voltage below V_DD.

Figure 4 shows the difference in how the 30-V UCC27624 and a 20-V V_DD gate driver react to high-frequency noise on the outputs.

Figure 4: Driver response to noise on the outputs

In this case, the 20-V V_DD driver experiences a logic error from the noise, whereas the higher V_DD of the UCC27624 eliminates errors without adding external components.

Conclusion

As electronics move forward, power supplies are going to keep moving to higher and higher frequencies in the quest to be more efficient, and transients will continue to grow with the frequency. When designing a power supply, remember that instead of adding extra components to your system to compensate, start with a driver that has the voltage range, negative voltage handling and reverse-current capability to address transients in your system.

Additional resources

• Read the application report, “Fundamentals of MOSFET and IGBT Gate Driver Circuits.”
• Check out the application note, “External Gate Resistor Design Guide for Gate Drivers.”

From smartphones to cars, consumers are demanding more features packed into increasingly smaller products. To help with this trend, TI has optimized its packaging technology for our semiconductor devices, including the load switches used for subsystem control and power sequencing. Packaging innovation enables greater power density, making it possible to fit even more semiconductor devices and features onto each printed circuit board.

Wafer chip-scale packaging (WCSP)

Today, the smallest-size load switches available are in wafer chip-scale packaging (WCSP). Figure 1 shows an example of a four-pin WCSP device. 

Figure 1: Four-pin WCSP device

The WCSP technique takes the silicon die and attaches solder balls to the bottom, giving it the smallest footprint possible and making the technology competitive in terms of current carrying and package area. Because WCSP minimizes form factor, the number of solder balls used for input and output pins limit the maximum current that the load switch can support.

Plastic packaging with wire-bond technology

Higher-current applications or those with harsher manufacturing processes such as industrial PCs require the implementation of plastic packaging. Figure 2 shows a traditional plastic package implementation with wire-bond technology.

Figure 2: Standard wire-bond quad-flat no lead (QFN) package

QFN or small-outline no lead (SON) packages use wire-bond technology to connect the die to the leads, which enables higher amounts of current to pass from input to output while providing good thermal characteristics for self-heating. Wire-bond plastic packages need a lot of room for the bond wires themselves, however, necessitating a bigger package when compared to the die size itself. Bond wires also add resistance to the power path, increasing the total on-resistance of the load switch. Here, the trade-off is between a larger size and higher power support.

 Plastic HotRod  package 

While both WCSP and wire-bond packaging have their advantages and limitations, TI’s HotRod  QFN load switches offer the advantages of both packaging technologies. Figure 3 shows a breakdown of a HotRod package.

Figure 3: TI HotRod QFN structure and die attachment

These leadless plastic packages use copper pillars to connect the die to the package, and since this takes less area than bond wires, it becomes possible to minimize the package size. The pillars also support high current levels and add minimal resistance to the power path, allowing as much as 6 A of current through a single pin.

Table 1 illustrates these advantages through a comparison of the TPS22964C WCSP, TPS22975 wire-bond SON and TPS22992 load switches.

Table 1: Comparison of various load switch solutions

While the TPS22975 wire-bond SON device can also support 6 A of current, achieving this current level requires two pins for both input and output voltages, which limits the number of additional features such as power good and adjustable rise time. The bond wires also add to the on-resistance of the device, limiting the maximum current.

The WCSP load switch is the smallest of the three solutions, but its limited pins give it the smallest number of features and the lowest current support.

Conclusion

The TPS22992 load switch combines the advantages of both WSCP and SON, offering the small-size advantage of the WCSP solution with the high current support and additional features of the wire-bond SON solution. TI’s TPS22992 and TPS22998 load switches use the HotRod package to optimize small solution size while supporting high current, low on-resistance and many device features.

Additional resources

• Read the technical white paper, “When to Make the Switch to an Integrated Load Switch.”
• Learn more about load switches in the application report, “Basics of Load Switches.”
• Explore design tips in the ebook, “11 Ways to Protect Your Power Path.”
• Read the application report, “Enhanced HotRod QFN Package: Achieving Low EMI Performance in Industry’s Smallest 4-A Converter.”
• Check out TI’s QFN and WCSP packaging solutions.

Gallium nitride (GaN) is a popular topic in the power electronics industry, as it enables designs such as 80 Plus titanium power supplies, 3.8-kW/L electric vehicle (EV) onboard chargers and EV charging stations. In many applications, GaN replaces traditional silicon metal-oxide-semiconductor field-effect transistors (MOSFETs) given its ability to drive higher power density and efficiency. But because of its electrical properties and the performance that it enables, designing with GaN has a different set of challenges than silicon.

Several different types of GaN FETs exist with different device structures – depletion mode (d-mode), enhancement mode (e-mode) and cascade to cathode (cascode) – and each has its own accompanying gate-driver and system requirements. In this article, I will break down the most important considerations for designing with different types of GaN FETs to improve power density in your system design. I will also review how integrating functions such as a gate driver and voltage supply regulation can significantly simplify your overall design.

Anatomy of a GaN FET

Each GaN power switch must be paired with an appropriate gate driver. (Otherwise, you could experience a pop and puff of smoke when testing at the bench!) GaN devices can have uniquely sensitive gates, as they are not classical MOSFETs, but are instead high-electron-mobility transistors (HEMTs). The cross section of a HEMT, shown in Figure 1, appears similar to a MOSFET; however, current does not flow through the full substrate or buffer layer, but instead flows through a two-dimensional electron gas layer.

Figure 1: Cross-section of the lateral structure of GaN FETs

Incorrect gate control of a GaN FET will cause a breakdown of the insulative layer, barrier or other structural elements; the device will not only fail during that system condition, but it is likely also permanently damaged. This level of sensitivity necessitates a review of different types of GaN devices and their broad needs. HEMTs also do not have the traditional doped FET structure that forms p-n junctions, which then cause body diodes. This means that there are no internal diodes that can break down or cause unwanted behavior during operation such as reverse recovery.

Gate-driver and bias-supply considerations

E-mode GaN FETs look very similar to the e-mode silicon FETs that you may already have experience with. A positive voltage of 1.5 V to 1.8 V will begin to turn on the FET, with most operating conditions specified for 6-V gate threshold operation. However, most e-mode GaN devices have a maximum gate threshold of 7 V, which when violated will likely result in permanent damage.

Because traditional silicon gate drivers may not offer proper voltage regulation or cannot handle the high common-mode transient immunity in a GaN-based design, many designers choose a gate driver such as the LMG1210-Q1, which TI designed specifically for use with a GaN FET. This device offers a gate-drive voltage of 5 V, regardless of supply voltage. Traditional gate drivers would need very tight regulation of the gate driver’s bias supply so that they do not overstress the GaN FET. A cascode GaN FET, shown in Figure 2, is a compromise for ease of use compared to e-mode GaN FETs.

Figure 2: Symbols for e-mode and cascode d-mode GaN FETs

The GaN FET is a d-mode device, which means that it is normally on, and requires a negative gate threshold to turn the device off. This is extremely problematic for a power switch, so most manufacturers add a 30-V silicon FET in series with the GaN FET for sale as one package. The gate of the GaN FET connects to the source of the silicon FET, and applies the turnon and turnoff gate pulses to the gate of the silicon FET.

The biggest benefit of this approach is that traditional isolated gate drivers such as the UCC5350-Q1 drive the silicon FET, eliminating many gate-driver and bias-supply concerns. The biggest downsides of cascode GaN FETs are higher output capacitance for the FET and susceptibility to reverse recovery given the presence of a body diode. The output capacitance of the silicon FET adds on to that of the GaN FET, resulting in a 20% increase, which means >20% increased switching losses compared to other GaN solutions. And during reverse conduction, the body diode of the silicon FET conducts current and undergoes reverse recovery when the voltage polarity flips.

Cascode GaN FETs operate at slew rates of 70 V/ns (compared to 150 V/ns for other GaN solutions) in order to guard against avalanche breakdown of the silicon FET, increasing switching overlap losses. Although cascode GaN FETs are simpler to design with, they limit the achievable performance.

Integration offers an easier solution

The integration of a gate driver with built-in bias-supply regulation and a d-mode GaN FET solves many of the design challenges of e-mode and cascode GaN FETs. For example, the LMG3522R030-Q1, a 650-V 30-mΩ GaN device, has an integrated gate driver and power management features that enables higher power density and efficiency while reducing the risks and engineering effort required. Since the GaN FET is d-mode, there is a silicon FET integrated in series with the GaN FET. But the big difference versus cascode GaN FETs is that the integrated gate driver can directly drive the gate of the GaN FET, while the silicon FET performs the role of a normally-off enable switch at power-up. This approach, known as direct drive, eliminates the most pressing issues of cascode GaN FETs, such as higher output capacitance, reverse-recovery susceptibility and avalanche breakdown of the silicon FET in series. The gate driver integrated in the LMG3522R030-Q1 enables very low switching overlap losses, enabling the GaN FET to operate at a switching frequency as high as 2.2 MHz and eliminating the risk of pairing a GaN FET with the wrong gate driver. Figure 3 shows an example of a half-bridge configuration using integrated LMG3522R030-Q1 GaN FETs.

Figure 3: Simplified GaN half-bridge configuration using the UCC25800-Q1 transformer driver and two LMG3522R030-Q1 GaN FETs

The integrated driver shrinks solution size, enabling a power-dense system. Integrating a buck-boost converter also means that the LMG3522R030-Q1 can operate with a 9-V to 18-V unregulated power supply, which significantly reduces bias-supply requirements. To enable a compact and lower-cost system solution, you could combine the LMG3522R030-Q1 with an ultra-low electromagnetic interference transformer driver such as the UCC25800-Q1, which has open-loop inductor-inductor-capacitor control with multiple secondary-side windings. Alternatively, a highly integrated, compact bias power supply such as UCC14240-Q1 DC/DC module can supply the device locally, resulting in a low-profile design with a small printed circuit board footprint.

Conclusion

With the right gate driver and bias supply, GaN devices can help you achieve system-level benefits such as a switching speed of 150 V/ns, reduced switching losses and a smaller magnetics size for high-power systems across industrial and automotive applications. Integrated GaN solutions simplify many of your device-level challenges so that you can focus on the wider system.

Additional resources

Check out these white papers:

• “Optimizing GaN Performance With an Integrated Driver.”
• “Direct Drive Configuration for GaN Devices.”
• “Understanding the Trade-Offs and Technologies to Increase Power Density.

View our reference designs:

• 3.6-kW single-phase totem-pole bridgeless PFC reference design with a > 180-W/in³ power density
• 6.6-kW three-phase, three-level ANPC inverter/PFC bidirectional power stage reference design

It’s a challenge to create highly efficient and compact designs while also adhering to strict electromagnetic interference (EMI) requirements imposed by groups such as Comité International Spécial des Perturbations Radioélectriques (CISPR). Therefore, component selection becomes a critical part of the design process. As with most design decisions, choosing between different components almost always comes down to an assessment of tradeoffs based on your most critical design goals. Known for high efficiency and good thermal performance, buck regulators are not typically considered low-EMI options. Fortunately, you have several options for reducing the EMI generated by such regulators. To aid further discussion, Figure 1 shows a simplified buck regulator schematic.

Figure 1: Simplified buck regulator schematic

Board layout considerations

Beyond selecting proper passive component values to ensure a functional design, board layout should be your first consideration when your design must fall under EMI limits. There are two general rules that can help minimize generated EMI with all buck regulator board layouts:

• Minimize high transient current (di/dt) loop areas by bringing the input capacitor and boot capacitor as close to the VIN and GND pins of the integrated circuit as possible.
• Minimize the surface area of high transient voltage (dv/dt) nodes by minimizing the area of the switch node.

In instances where board layout optimization is not possible, there are other options. Learn more about them in the technical article, “How Device-Level Features and Package Options Can Help Minimize EMI in Automotive Designs.”

Integrated input capacitors

As I mentioned, reducing the area of high di/dt current loops is very important when designing switching regulators to remain under EMI limits. In a buck regulator, it’s important to consider the input-voltage-to-ground loop from an EMI perspective. A buck regulator steps down a higher DC voltage to a lower one by switching the connection to the supply on and off, resulting in high-side metal-oxide semiconductor field-effect transistor (MOSFET) (Q1) current, shown in Figure 2.

Figure 2: Input current waveform to a buck regulator

The MOSFET switches on and off rapidly, creating very sharp, almost discontinuous currents supplied by the input capacitor. Some devices, such as TI’s 3-A LMQ66430-Q1 and 6-A LMQ61460-Q1 36-V buck regulators, integrate high-frequency input capacitors inside the package, resulting in the smallest possible input current-loop area. Reducing the area of this input current loop results in smaller parasitic inductance at the input, which reduces the amount of electromagnetic energy emitted.

Integrated boot capacitor

Another high di/dt current loop that you should consider is the boot capacitor loop. The boot capacitor is responsible for supplying charge to the high-side MOSFET gate driver during the on-time. Internal circuitry refreshes this capacitor during the off-time. The source terminal of the high-side MOSFET connects to the switch node rather than GND. Referencing the boot capacitor to the source pin of the MOSFET ensures that the gate-to-source voltage (V_GS) is high enough to turn on the MOSFET. With most buck regulators, you will have to leave some switch node area available on the board to connect the bootstrap capacitor, although this can be counterproductive when trying to minimize the area of the switch node for EMI. By integrating the boot capacitor inside the package, the LMQ66430-Q1 follows the two rules that I mentioned earlier, while also reducing the need for an external component.

Conclusion

It can be difficult to design compact power-supply designs capable of remaining under strict EMI limits. Buck regulators with integrated capacitors can make the process of EMI-compliant designs easier, while also helping reduce the overall external component count.

Additional resources

Check out these technical articles:

• “How a DC/DC Converter Package and Pinout Design Can Enhance Automotive EMI Performance.”
• “Lowering Audible Noise in Automotive Applications with TI’s DRSS Technology.”
• For a comprehensive overview of all things related to low EMI, read the e-book, “An Engineer’s Guide to Low EMI in DC/DC Regulators.”

The adoption of gallium nitride (GaN) field-effect transistor (FET) is rapidly increasing, given its ability to improve efficiency and shrink the size of power supplies. But before investing in the technology, you may still be asking yourself whether GaN is reliable. It strikes me that no one asks whether silicon is reliable. After all, there are still new silicon products coming out all the time, and power-supply designers also care about the reliability of silicon power devices.

The truth is, the GaN industry has invested considerable effort and time on reliability.

For silicon, the reliability question is phrased differently – “Did it pass qualification?” Although GaN parts also pass silicon qualification, power-supply manufacturers are not convinced that the silicon methodology assures reliable GaN FETs. This is a valid point, since not all silicon device tests are applicable to GaN, and traditional silicon qualification itself does not include stress tests for the actual switching conditions of power-supply usage.  The JEDEC JC-70 Wide Bandgap Power Electronic Conversion Semiconductors committee has released several GaN-specific guidelines to address the deficiencies.

Understanding GaN Product Reliability

Read the technical white paper, “Achieving GaN Products with Lifetime Reliability,” to learn more about our GaN reliability testing.

How do you validate GaN’s reliability?

GaN FET reliability is validated through the established silicon methodology, combined with reliability procedures and test methods designed to address GaN-specific failure modes such as an increase in the dynamic drain-to-source on-resistance (R_DS(ON)). Figure 1 lists the steps to achieving reliable GaN products.

Figure 1:  GaN-specific guidelines for reliability in conjunction with established silicon standards

We have categorized testing into component- and power-supply-level blocks, with relevant standards and guidelines for each block. At the component level, TI runs bias, temperature and humidity stress tests according to the traditional silicon standards, uses GaN-specific test methods and determines lifetimes by applying accelerated stress until device failure. At the power-supply level, parts are run under stringent operating conditions for the relevant application. TI also validates robustness for extreme operating conditions from occasional events.

The reliability of GaN FETs in your application

The JEDEC JEP180 guideline provides a common approach for assuring the reliability of GaN products in power-conversion applications. To meet JEP180, a GaN manufacturer must demonstrate that its product has the required switching lifetime for the relevant stress and runs reliably under stringent operating conditions in a power supply. The former demonstration uses switching accelerated lifetime testing (SALT) to stress devices to failure, and the latter uses dynamic high-temperature operating life (DHTOL) testing.

Devices are also subject to extreme operating conditions in the real world – events such as short circuits and power-line surges. TI GaN products such as the LMG3522R030-Q1 have built-in short-circuit protection. Surge robustness across a range of applications entails the consideration of both hard- and soft-switching stresses. GaN FETs handle power line surges differently from silicon FETs. Rather than enter avalanche breakdown, GaN FETs switch through surge strikes as a result of their overvoltage ability.  An overvoltage ability can also improve system reliability because avalanching FETs cannot absorb much avalanche energy, so the protection circuitry must absorb most of the surge. The degradation of surge-absorption components with aging subjects silicon FETs to higher levels of avalanche, which can cause failures. In contrast, a GaN FET will continue switching.

Are TI GaN products reliable?

TI qualifies its GaN products according to the methodology shown in Figure 1. Figure 2 summarizes the outcomes, showing the results from both component- and power-supply-level blocks.

Figure 2: GaN FET reliability is validated by GaN specific guidelines using the methodology shown in Figure 1.

At the component level, TI GaN passes traditional silicon qualification and has high reliability for GaN-specific failure mechanisms. TI has engineered and validated high reliability for time-dependent breakdown (TDB), charge trapping and hot-electron wearout failure mechanisms, and demonstrated that dynamic R_DS(ON) is stable with aging.

 To determine the component switching lifetime, our SALT validation applies accelerated hard-switching stress, as described in “A Generalized Approach to Determine the Switching Lifetime of a GaN FET.” The TI model uses switching waveforms to directly calculate the switching lifetime, and shows that a TI GaN FET will not fail from hard-switching stresses throughout the product’s lifetime.

 To validate power-supply level reliability, we ran DHTOL testing on 64 TI GaN parts under stringent power-supply usage conditions. Devices showed stable efficiency with no hard failure, demonstrating reliable operation for all modes of power-supply operation: hard and soft switching, third-quadrant operation, hard commutation (reverse recovery), Miller shoot-through with high slew rates, and reliable interaction with the driver and other system components. TI has additionally validated surge robustness by applying surge strikes to devices running in power supplies under both hard- and soft-switching operation, and showed that TI GaN FETs can effectively switch through bus voltage surges as high as 720 V, which provides significant margin. Learn more about this testing in “A New Approach to Validate GaN FET Reliability to Power-Line Surges Under Use Conditions.”

 Conclusion

The GaN industry has built a methodology to assure the reliability of GaN products, so the question is not “Is GaN reliable?”, but “How do you validate GaN’s reliability?” TI GaN devices are reliable at both the component level and in real-world applications. They have passed both silicon qualification standards and GaN industry guidelines. In particular, TI GaN products pass JEP180, demonstrating that they are reliable for power-supply usage.

Additional resources

• Browse TI’s line of GaN products.
• Learn about the benefits of GaN technology.

Relays have been used as switches since before the transistor was invented. The ability to safely control high-voltage systems from lower-voltage signals, as is the case in isolation resistance monitoring, is necessary for the development of many automotive systems. While the technology of electromechanical relays and contactors has improved over the years, it is still challenging for designers to achieve their goals of lifetime reliability and fast switching speeds, along with low noise, shock vibration and power consumption.

Solid-state relays (SSRs) exhibit performance and cost benefits and are rated for different levels of isolation, whether basic or reinforced. SSRs also possess advantages over alternative technologies such as electromechanical relays and solid-state photorelays.

Achieve higher reliability in your system

Learn more about our solid-state relay portfolio.

Traditional relay switching solutions

Electromechanical relays (EMRs) are common in high-voltage switching applications. EMRs employ the use of electromagnetic forces to mechanically switch contacts on and off. Given their mechanical nature, EMRs feature an incredibly low on-resistance; their contacts are essentially a metal-to-metal connection.

EMRs do have trade-offs when it comes to switching speeds and reliability. Moving parts inside the relay are a limiting factor, and switching speed is typically in the 5- to 15-ms range. Over time and with use, an EMR can experience failures such as arching, chattering and welding shut.

Unlike EMRs, photorelays have no moving parts and provide a high isolation voltage. Photorelays are an improvement over traditional EMRs; however, they also have design considerations such as limitations on the achievable power transfer as well as deterioration of the internal LED. Additionally, photorelays need an external current-limiting resistor and often use additional field-effect transistors (FETs) to manage the LED’s on or off state.

Higher-reliability isolation using SSRs

Solid-state relays from TI are available as switches (with integrated FETs) or drivers for controlling external FETs. Whether leveraging capacitive or magnetic isolation, TI’s isolated SSR portfolio can enable a design with basic or reinforced levels of isolation. TI’s TPSI2140-Q1 isolated switch and TPSI3050-Q1 isolated driver feature higher reliability and longevity compared to EMRs, since they do not experience mechanical deterioration over time. SSRs thus enable a 10 times higher lifetime reliability than traditional EMRs. TI’s SSRs can also switch in the microsecond range, orders of magnitude faster than EMRs.

Since the TPSI3050-Q1 and TPSI2140-Q1 integrate power and signal transfer across a single isolation barrier, no secondary bias supply is necessary, making it possible to achieve a small solution size. Figure 1 illustrates the use of the TPSI2140-Q1 isolated switch in a high-voltage system, eliminating external components such as a bias supply and external control circuits.

Figure 1: The TPSI2140-Q1 isolated switch reduces solution size in high-voltage systems

Solid-state relays such as the TPSI2140-Q1 and TPSI3050-Q1 also offer advantages over traditional photorelays and optocouplers. TI devices such as the TPSI2140-Q1 and TPSI3050-Q1 achieve better reliability over photorelays because there is no LED degradation. No external control circuits are necessary because the logic-level input can drive the system directly. Table 1 qualitatively compares these isolated switching technologies.

Table 1: Comparing existing switching solutions

Conclusion

TI’s solid-state relays provide the highest dielectric strength at the fastest speed, highest operating temperature and lowest system cost. They also enable more reliable switching in a smaller package. Learn more about how we can help you simplify your design with our isolated switches and drivers in the additional resources below.

Additional resources

• View our introduction to isolation trainings in our TI Precision Labs training series.
• Read our application notes:
  • Why Pre-Charge Circuits Are Necessary in High Voltage Systems
  • Cascoding Two TPSI3050 Isolated Switch Drivers to Increase Gate Drive Voltage
• Evaluate the products with the TPSI3050Q1EVM and TPSI2140Q1EVM evaluation modules.

High-voltage industrial and automotive systems such as factory automation equipment, grid infrastructure applications, motor drives and electric vehicles (EVs) can generate several hundred to thousands of volts, which poses a significant safety risk to humans and can reduce equipment lifetimes. This article explains how to keep these HV systems safe by utilizing the latest isolation technologies for improved reliability while reducing solution size and cost.

Methods of isolation

Integrated circuits (ICs) achieve isolation by blocking DC and low-frequency AC currents while allowing power, analog signals or high-speed digital signals to transfer across the isolation barrier. Figure 1 shows three popular semiconductor technologies used to achieve isolation: optical (optocoupler), electric field signal transfer (capacitive) and magnetic field coupling (transformer).

(a)

(b)

(c)

Figure 1. Semiconductor isolation technologies: optocoupler (a); capacitive (b); transformer (c).

TI leverages capacitive isolation technology and proprietary integrated planar transformers (magnetic isolation), along with advanced packaging and process technology, to achieve some of the highest levels of reliability, integration and performance available with our large and diverse isolated-ICs portfolio.

Overcome high-voltage design challenges with reliable isolation technologies

Read our white paper to learn about common high-voltage galvanic isolation concerns and methods, and how to achieve high-voltage isolation reliably in industrial and automotive systems while reducing solution size and cost.

Capacitive isolation

Capacitive isolation technology is based on AC signal transfer across a dielectric. TI’s capacitive isolators are constructed using an SiO₂ dielectric, which provides very high dielectric strength. Since SiO₂ is an inorganic material, it’s also extremely stable over moisture and temperature. Further, our proprietary methodology for multilayered capacitor and multilayer passivation improves isolator quality and reliability by reducing the dependence of high-voltage performance on any single layer. Our capacitive technology supports working voltages (V_IOWM) of 2 kV_RMS, withstands isolation voltages (V_ISO) of 7.5 kV_RMS and has a surge voltage capability of 12.8 kV_PK.

Magnetic isolation

Magnetic isolation is typically used in applications that require high-frequency DC/DC power conversion. One advantage of IC transformer-coupled isolation is the ability to transfer power in excess of hundreds of milliwatts, which often eliminates the need for a secondary-side bias supply. It’s also possible to use magnetic isolation to send high-frequency signals. In systems that need to send both power and data, you can use the same transformer winding coils for power and the signal needs, as shown in Figure 2. By combining both signals and power transfer over the same integrated transformer coil, both solution cost and size are minimized. The TPSI3050-Q1 and TPSI3052-Q1 utilize combined data and power transfer over the same transformer channel.

Figure 2. Using magnetic isolation to send both power and signals reliably across an isolation barrier

TI uses a proprietary multichip module approach for magnetic isolation, which co-packages a high-performance planar transformer with an isolated power stage and dedicated controller die. We can build these transformers with either a high-performance ferrite core to improve coupling and transformer efficiency, or an air-core to save cost and complexity when the application requires only modest power transfer. 

Achieve isolation needs reliably while reducing solution size and cost.

Different applications require different isolation approaches. Let’s look at a few examples of how TI ICs can help solve high-voltage isolation needs in EV and grid infrastructure applications with very high reliability while also reducing solution size and cost.

EV applications

Reduced weight, increased torque, higher efficiency and faster charging are boosting high-voltage battery stacks in EVs from 400 V to levels of 800 V – even as high as 1 kV. Battery management systems (BMSs) and traction inverters are two of the most critical EV subsystems where the 800-V domain needs to be isolated from the chassis to help ensure the safety of passengers and their vehicles.

The block diagram shown in Figure 3 is an example of a traction inverter that uses isolated gate drivers to drive high-voltage insulated-gate bipolar transistor (IGBT) or silicon carbide (SiC) modules in a three-phase, DC-to-AC inverter configuration. These modules can co-package as many as six IGBT or SiC switches, requiring up to six isolation transformers, powering six independent gate-driver ICs. Our UCC14240-Q1 is a dual-output, mid-voltage, isolated DC/DC power module that can enable higher performance in traction inverter, gate-driver bias applications while minimizing PCB area by reducing the number of external transformers.

Figure 3. A typical traction inverter system block diagram.

Additionally, BMSs use a pre-charge circuit when connecting high-voltage battery terminals to subsystems. Our 5-kV_RMSTPSI3050-Q1 isolated switch driver was designed to replace mechanical pre-charge contactors to achieve a smaller, more reliable solid-state solution. It provides reinforced isolation up to 5 kV_RMS, and a 10 times higher operating lifetime than electromechanical relays, which degrade over time.  Figure 4 illustrates the area savings that the TPSI3050-Q1 can provide as compared to a mechanical relay.

Figure 4. Reduction of solution size using magnetic isolation based solid-state relay driver (TPSI3050).

Grid infrastructure applications

Isolation is essential in grid infrastructure applications to provide protection from high-voltage surges that may damage equipment or harm humans, eliminate disruptive ground loops in interconnections involving large ground potential differences (GPDs) and maintain data integrity during common-mode transient events.

Solar energy equipment and EV chargers can operate at voltages ranging from 200 V to 1,500 V or more. Figure 4 shows our AFE for Insulation Monitoring in High-Voltage EV Charging and Solar Energy Reference Design, reference design, which provides insulation resistance monitoring in grid infrastructure applications using our AMC3330 precision isolated amplifier and TPSI2140-Q1 isolated switch. Because there are no moving parts, this solid-state relay solution can perform frequent measurements for decades without any performance degradation. Both power and signals are transferable across the isolation barrier within the TPSI2140-Q1, so there’s no need for secondary-side bias supplies. And because the device is available in low-profile small outline IC (SOIC) packages, the solution size can be as much as 50% smaller than photorelay- or mechanical relay-based solutions.

Figure 5. AFE for insulation monitoring in high-voltage EV charging and solar energy block diagram.

Conclusion

By integrating more functionality within our isolation technology, TI is enabling engineers to maintain safety in applications like EVs and grid infrastructure while reducing design complexity, solution size and cost. See ti.com/isolationtechnology to see how we’re scaling our capacitive and magnetic isolation technologies to add more analog functions.

Additional resources

• Read the white paper, “Enabling High Voltage Signal Isolation Quality and Reliability,” for more information about the reliability of TI’s high-voltage isolation capacitors.
• Check out the application brief, “How to Simplify Isolated 24-V PLC Digital Input Module Designs.”
• Learn about essential isolation parameters, certifications, and how to design and troubleshoot with various types of isolated ICs by watching the TI Precision Labs – Isolation training series.

As demand for data increases, so does demand for servers and data centers, and thus higher demand for power. Industry trends suggest that power per rack, which was 4 kW in 2020, will be as high as 20 kW in 2025.

Given limited physical real estate available for data centers and servers, the delivery of more power in less area is known as a high power density requirement in server power architectures. Increasing the efficiency of server power supplies can also keep cooling costs down.

Everything around us is getting data-hungry and data-driven. All of this data is stored and processed by servers in data centers, as shown in Figure 1.

Figure 1: A data-connected ecosystem

Servers are usually scalable and are hot-swappable in order to meet different processing requirements and maintain high system availability. To achieve seamless hot-swap functionality, server motherboards and power distribution boards employ hot-swap controllers or eFuses. Components such as eFuses in server power supplies need to provide higher current to meet increased server power requirements. Protection devices such as hot swaps and eFuses also need to handle high peak current to match the higher peak-processing capabilities of modern microprocessors in servers. Figure 2 shows a typical server power architecture.

Figure 2: A typical server power architecture

Traditionally, high-power server designs include hot-swap controllers with multiple metal-oxide semiconductor field-effect transistors (MOSFETs). But server power and power-density requirements are increasing exponentially. To satisfy these needs and simplify these designs, consider the TPS25985 (80 A peak) and TPS25990 (60 A peak with the PMBus interface) eFuses in server power architectures. The TPS25985 and TPS25990 can support 60 A_DC and 50 A_DC, respectively, and have an adjustable current limit of up to 60 A and 50 A, also respectively. It is possible to stack multiple unlimited TPS25985 and TPS25990 eFuses to achieve higher current.

Achieving high power density

Power density is a must-have requirement for modern server power supply units (PSUs). The latest generation of server PSUs are in the range of a 3-kW (250 A at 12 V) power rating. When selecting an eFuse, it is important to have the highest current in the smallest size. The TPS25985 packs 80 A of peak current in a 4.5-mm-by-5-mm package. Figure 3 shows some of the TI’s eFuses.

Figure 3: Power-density progression of TI eFuses

By integrating a MOSFET, a current monitor, a comparator, active current sharing and a temperature monitor, the TPS25985 and TPS25990 eFuses significantly reduce the total printed circuit board or printed wiring board area. When connecting multiple TPS25985 and TPS25990 eFuses, the board savings and power density improve multifold. Figure 4 shows the current density of the TPS25985 and TPS25990 compared to other eFuses on the market.

Figure 4: Current and power-density comparison

Current-share and current-monitor accuracy

Hot-swap controllers cannot control the gates of multiple paralleled MOSFETs very precisely; therefore, current sharing by paralleled MOSFETs is not accurate. Precision amplifiers can help achieve high current-share accuracy and current-monitor accuracy, but adding them increases the total solution size. It is challenging to gauge the die temperature of the MOSFET, and therefore impossible to guarantee its thermal protection in transient and steady-state conditions. Figure 5 highlights key pins and functions of the TPS25985.

Figure 5: TPS25985 pinout, highlighting key differentiating features

The TPS25985 and TPS25990 eFuses have integrated active current sharing and direct access to MOSFET die parameters (voltage, current, temperature), which allows accurate control of all eFuse gates connected in parallel and accurate die temperature monitoring of integrated FETs. Compared to an eFuse without active current sharing, the TPS25985 and TPS25990 enable design engineers to optimize the number of eFuses and the performance of the system.

An integrated current monitor enhances server platform power management using PSYS/PROCHOT to maximize the platform’s computational throughput and power-supply utilization. These features also help optimize the front-end AC-to-DC power supply, and thus system cost. In addition, an adjustable transient current blanking timer improves system reliability and total availability by avoiding nuisance tripping.

Remote monitoring and control

The TPS25990 adds PMBus interface capability to the system. The TPS25990 enables single-command power cycling with an adjustable turnon delay, which allows system design engineers to sequence and reset the system remotely. The TPS25990 also offers black-box capability, where seven events are recorded with relative timestamps. The TPS25990 incorporates high-speed analog-to-digital converters that enable users to plot one signal of their choice, mimicking a digital oscilloscope. The GUI for the TPS25990, along with its other features, helps design engineers not only reduce their total development time but also quickly identify and resolve field issues, which are generally very difficult to reproduce and troubleshoot.

Thermal considerations

Server power systems operate at a wide ambient temperate range (–40ºC to 85ºC). Hot-swap controllers or eFuses experience even higher ambient temperatures. Therefore, power design engineers become concerned about the thermal performance of these devices when high currents are packed in small packages. The TPS25985 and TPS25990 eFuses alleviate this concern, with the ability to operate at a 125ºC junction temperature. The TPS25985 and TPS25990 offer an R_DS(on) of 0.59 mΩ and 0.79 mΩ, respectively; R_DS(on) spread across process, voltage and temperature variations is limited. Thus, these eFuses experience very low self-heating and a wide operating temperature range without sacrificing the derating. Figure 6 shows the case temperature of the TPS25985.

Figure 6: TPS25985 case temperature at V_IN = 12 V, I_OUT = 50 A, Tamb = 25°C

Conclusion

Design engineers can reduce development time and design cost by using the TPS25985 and TPS25990 eFuses in server power architectures. The eFuses’ low R_DS(on) reduces power losses in the system, helping data centers achieve their efficiency goals. Improved power density improves data center processing capabilities and gives end users a seamless experience across their data-connected devices. Better diagnostic, scaling and configuration capabilities in eFuses can help data centers minimize downtime, which helps maintain continuity of service and their ability to offer customers high-percentage uptime guarantees.

The number of semiconductors in nearly every application is multiplying, and many of the design challenges facing electronics engineers all tie back to the need for greater power density. A few example applications come to mind:

• Hyperscale data centers: Rack servers are using an incredible amount of power, challenging utility companies and power engineers to keep up with increasing demands.
• Electric vehicles: The transition from internal combustion engines to 800-V battery packs comes with an exponential increase in semiconductor content for the powertrain.
• Commercial and home security applications: As video doorbells and Internet Protocol cameras become more prevalent, their shrinking sizes put constraints on the necessary thermal solution.

Understanding the Trade-offs and Technologies to Increase Power Density

Thermal efficiency is paramount to improving power density. Read the white paper to learn how to overcome common barriers.

What stands in the way of achieving higher power density? Well, thermal performance is an electrical byproduct of power-management integrated circuits (ICs), which you can’t ignore or “optimize out” with filtering components at the system level. The mitigation of thermals requires critical microadjustments throughout every step in the development process so that the design can achieve its system requirements for a given size constraint. Following are three key areas TI focuses on to optimize thermal performance and break through power-density barriers at the chip level.

1. Process technology innovations

Many global semiconductor manufacturers are racing to offer power-management products that leverage process technology nodes to achieve higher performance capabilities in industry-standard packages. For example, at TI we continue investing in 45- and 65-nm process technologies that leverage our internal technology development, along with 300-mm manufacturing efficiencies to offer products optimized for cost, performance, power, precision and voltage levels. Our process technology advancements also help us create products that maintain high performance under various thermal conditions. For instance, reducing the specific on-state resistance (R_SP) or drain-to-source on-state resistance (R_DS(on)) of integrated metal-oxide semiconductor field-effect transistors (MOSFETs) minimizes die size while enhancing thermal performance. The same is true for other semiconductor switches such as gallium nitride (GaN) or silicon carbide.

Take the TPS566242 buck converter, shown in Figure 1. New process nodes optimize the pin layout by integrating features and providing an extra ground connection that helps deliver 6 A of output current from a 1.6-mm-by-1.6-mm small outline transistor (SOT)-563 package. If you asked me five years ago if tiny, simple leaded packages would be capable of that type of performance, I’d have been skeptical. But that’s the beauty of process technology!

Figure 1: The TPS566242 synchronous buck converter delivers up to 6 A of continuous current

2. Circuit-design techniques

In addition to more efficiency at the process technology level, creative circuit design also plays an important role in improving power density. Designers have historically used discrete hot-swap controllers to protect high-current enterprise applications. As a protection function, they are reliable, but with end-equipment manufacturers (and consumers) requiring more current capability, discrete power designs can grow far too large, especially for applications such as server power-supply units (PSUs) that often require 300 A of current or more.

The TPS25985 eFuse pairs an integrated 0.59-mΩ FET with a current-sense amplifier. This amplifier, plus a new active current sharing approach, provides an easy way to allow for temperature monitoring. By pairing efficient switches with creative integration approaches, the TPS25985 can deliver up to 70 A of peak current, and you can easily stack multiple eFuses for higher power.

3.Thermally optimized packaging R&D

Although reducing the amount of heat dissipated into the printed circuit board (PCB) or system is is a basic requirement, the reality is that undesirable heat still lingers, especially as the power requirements or your system’s ambient temperature increase. TI has recently enhanced the performance of its HotRod quad-flat-no lead (QFN) packages, including larger die-attach pads (DAPs) to facilitate greater heat dissipation. Figure 2 shows the total DAP area and pin accessibility of the 6-A, 36-V TLVM13660 step-down power module.

Figure 2: The TLVM13660 includes four thermal pads on the bottom and has all signal and power pins accessible from the perimeter for easier layout and handling

To learn more about the evolution of these packages, see the Analog Design Journal article, “Designing with small DC/DC converters: HotRod QFN vs. Enhanced HotRod QFN packaging.”

System-level thermal solutions

For high-power applications like server PSUs, GaN with top-side cooling is a highly effective way to remove heat from the IC without heating up the PCB. The LMG3522R030-Q1 GaN FET integrates a gate driver and protection features in a top-side cooled package. Figure 3 illustrates the isolated DC/DC section of the 3-kW Phase-Shifted Full Bridge With Active Clamp Reference Design with >270-W/in³ Power Density, which leverages the LMG3522 and achieves a peak efficiency of 97.74%.

Figure 3: 3-kW phase-shifted full bridge with active clamp reference design

Of course, you may wish to have flexible cooling options, given variables such as the number of layers in your PCB or your assembly processes and system cost constraints. In those scenarios, bottom-side cooled ICs such as the LMG3422R030 integrated GaN FET may be more suitable.

Conclusion

Maintaining performance while reducing the impact of thermals can only happen with a multifaceted approach to process and packaging technology and power design expertise. At TI, it’s a challenge that our product designers, systems engineers, packaging R&D and manufacturing teams are all hyperfocused on – to achieve greater power density without thermal pitfalls.

Additional resources

• Find devices and technical resources to solve your density challenges at ti.com/powerdensity.
• Watch our training series, “Understanding the Fundamental Technologies of Power Density.”
• Create a custom power-supply circuit with WEBENCHâ Power Designer.