Home »Market News» How Smart Hardware Engineers Easily Design a Power Supply: Mini Tutorial

This article provides a simplified understanding and re-appreciation of the art of power supply design.

In the process of continuously reducing power consumption in more and more complex technologies and applications, what has happened in the field of power management? What about applications that deal with higher and higher voltages? This month's In Focus focused on various design development and manufacturing strategies taking place in the field of power management.

Most electronic systems require some kind of voltage conversion between the power supply voltage and the circuit voltage that needs to be powered. When the battery loses charge, the voltage drops. Some DC-DC conversion can ensure that more energy stored in the battery is used to power the circuit. In addition, for example, for 110 V AC lines, we cannot directly power semiconductors such as microcontrollers. Since voltage converters (also called power supplies) are used in almost all electronic systems, they have been optimized for different purposes over the years. Of course, some common optimization goals are solution size, conversion efficiency, EMI, and cost.

The simplest power supply: LDO

One of the simplest forms of power supply is a low dropout (LDO) regulator. The LDO is a linear regulator, not a switching regulator. The linear regulator places an adjustable resistor between the input voltage and the output voltage, which means that the output voltage is fixed, regardless of how the input voltage changes and the load current through the device. Figure 1 shows the basic principle of this simple voltage converter.

Figure 1: A linear regulator converts one voltage to another.

For many years, a typical power converter consisted of a 50 Hz or 60 Hz transformer connected to the grid, with a certain winding ratio to generate an unregulated output voltage, several volts higher than the power supply voltage required in the system. Then, according to the needs of the electronic equipment, a linear regulator is used to convert this voltage into a well-regulated voltage. Figure 2 shows a block diagram of this concept.

Figure 2: The linear voltage regulator after the line transformer.

The problem with the basic setup in Figure 2 is that the 50 Hz/60 Hz transformer is relatively bulky and expensive. In addition, the linear regulator will dissipate a lot of heat, so the overall system efficiency is low, and it is difficult to eliminate the heat generated when the system power is high.

Switch mode power supply for rescue

In order to avoid the shortcomings of the power supply shown in Figure 2, a switch-mode power supply was invented. They do not rely on 50 Hz or 60 Hz AC voltage. They use DC voltage, sometimes rectified AC voltage, and generate a higher frequency AC voltage to use a smaller transformer, or in a non-isolated system, use an LC filter to rectify the voltage to generate a DC output voltage. The advantage is that the solution size is small and the cost is relatively low. The AC voltage generated does not need to be a sinusoidal voltage waveform. A simple PWM signal shape can work normally, and it is easy to generate with a PWM generator and switch.

Until 2000, bipolar transistors were the most commonly used switches. They work well, but the switching speed is relatively slow. Their power efficiency is not high, limiting the switching frequency to 50 kHz or 100 kHz. Today, we use switching MOSFETs instead of bipolar transistors to achieve faster switching. This in turn provides us with lower switching losses, allowing switching frequencies up to 5 MHz. Such a high switching frequency allows the use of very small inductors and capacitors in the power stage.

Switching regulators bring many benefits. They usually provide power-efficient voltage conversion, allow voltage step-up and step-down, and provide a relatively compact and low-cost design. The disadvantage is that their design and optimization are not so simple, and they will generate EMI from the switching and switching frequency. The emergence of switch-mode power regulators and power design tools such as LTpowerCAD and LTspice have greatly simplified this difficult design process. Using these tools, the circuit design process of switching power supplies can be semi-automated.

When designing a power supply, the first question to be answered is whether galvanic isolation is required. Galvanic isolation is used for many reasons. It can make circuits safer, allow floating system operation, and prevent noisy ground currents from spreading through different electronic devices in a circuit. The two most common isolation topologies are flyback and forward converters. However, in order to obtain higher power, other isolated topologies can be used, such as push-pull, half-bridge, and full-bridge.

If galvanic isolation is not required, use a non-isolated topology in most cases. Isolated topologies always require transformers, and such devices are often expensive, bulky, and often difficult to meet the exact requirements of custom power supplies.

The most common topology when isolation is not required

The most common non-isolated switch-mode power supply topology is a buck converter. It is also called a buck converter. It accepts a positive input voltage and produces an output voltage lower than the input voltage. It is one of the three most basic switch-mode power supply topologies, requiring only two switches, an inductor, and two capacitors. Figure 3 shows the basic principle of this topology. The high-side switch inputs a pulse current and generates a switch node voltage that alternates between the input voltage and the ground voltage. The LC filter obtains the pulse voltage on the switch node and generates a DC output voltage. According to the duty cycle of the PWM signal controlling the high-side switch, DC output voltages of different levels are generated. This DC-DC step-down converter is very energy-efficient, relatively easy to construct, and requires few components.

Figure 3: The concept of a simple buck converter.

The buck converter produces pulsed current on the input side, and continuous current from the inductor on the output side. This is the reason why the buck regulator has a lot of noise on the input side and not much noise on the output side. It is important to understand this when you need to design a low-noise system.

In addition to the buck topology, the second basic topology is the boost or boost topology. It uses the same five basic power components as the buck converter, but rearranged so that the inductor is placed on the input side and the high-side switch is placed on the output side. The boost topology is used to boost an input voltage to an output voltage higher than the input voltage.

Figure 4: The concept of a simple boost boost converter.

When choosing a boost converter, it is important to note that the boost converter always specifies the maximum rated switching current instead of the maximum output current in its data sheet. In a buck converter, the maximum switching current is directly related to the maximum achievable output current, and has nothing to do with the voltage ratio between the input voltage and the output voltage. In the boost regulator, the voltage ratio directly affects the maximum possible output current according to the fixed maximum switching current. When choosing the right boost regulator IC, you not only need to know the required output current, but also the input and output voltages designed in development.

The input side noise of the boost converter is very low, because the inductance consistent with the input connection prevents the current from changing rapidly. However, on the output side, this topology is very noisy. We only see the pulse current flowing through the external switch, so the output ripple is more concerned than the buck topology.

The third basic topology consists of only five basic components and is an inverting buck-boost converter. The name stems from the fact that the converter takes a positive input voltage and converts it to a negative output voltage. In addition, the input voltage may be higher or lower than the absolute value of the inverted output voltage. For example, 5 V or 24 V at the input may produce a –12 V output voltage. This is possible without any special circuit modification. Figure 5 shows the circuit concept of the inverting buck-boost converter.

Figure 5: The concept of a simple inverting buck-boost converter.

In the inverting buck-boost topology, the inductor is connected from the switch node to ground. Both the input and output of the converter see pulse currents, making this topology relatively noisy at both the input and output. In low-noise applications, this characteristic can be compensated by adding additional input and output filtering.

A very positive aspect of the inverting buck-boost topology is that any buck switching regulator IC can be used in this type of converter. It is as simple as connecting the output voltage of the step-down circuit to the system ground. The ground of the step-down IC circuit will become the adjusted negative voltage. This characteristic has produced a lot of choices in the switching regulator ICs on the market.

In addition to the three basic non-isolated switch-mode power supply topologies discussed earlier, there are more topologies available. However, they all require additional power supply components. This usually makes them more costly and lower power conversion efficiency. Although there are some exceptions, adding additional components to the power path usually increases losses. Some of the most popular topologies are SEPIC, Zeta, Ćuk, and 4-switch buck-boost. They each provide functions not provided by the three basic topologies. The following is a list of the most important features of each topology:

SEPIC can generate a positive output voltage from a positive input voltage that may be higher or lower than the output voltage. The boost regulator IC can be used to design SEPIC power supplies. The disadvantage of this topology is that it requires a second inductor or a coupled inductor and a SEPIC capacitor.

Zeta converter is similar to SEPIC, but it can produce positive or negative output voltage. In addition, it has no right half plane zero (RHPZ), which simplifies the regulation loop. Buck converter ICs can be used for this type of topology.

The Ćuk converter can convert a positive input voltage to a negative output voltage. It uses two inductors, one on the input side and one on the output side, so the noise on the input and output sides is very low. The disadvantage is that there are not many switch-mode power conversion ICs that support this topology, because the regulation loop requires a negative voltage feedback pin.

This type of converter has become very popular in recent years. It provides a positive output voltage from a positive input voltage. The input voltage may be higher or lower than the adjusted output voltage. This converter replaces many SEPIC designs because it provides higher power conversion efficiency and only requires an inductor.

In addition to non-isolated topologies, some applications also require galvanically isolated power converters. The reason may be a safety issue, the need to use floating ground in larger systems interconnected with different circuits, or to prevent ground current loops in noise-sensitive applications. The most common isolated converter topologies are flyback and forward converters.

Flyback converters are usually used for power levels up to 60 W. The way this circuit works is that during the on-time, energy is stored in the transformer. During the shutdown period, this energy is released to the secondary side of the converter to power the output. The converter is easy to construct, but it requires a relatively large transformer to store all the energy required for normal operation. This aspect limits the topology to lower power levels. Figure 6 shows the flyback converter at the top and the forward converter at the bottom.

Figure 6: Flyback converter (top) and forward converter (bottom).

In addition to flyback converters, forward converters are also very popular. It uses a transformer in a different way than the flyback type. During the conduction period, although current flows through the primary winding, current also flows through the secondary winding. Energy should not be stored in the iron core of the transformer. After each switching cycle, we must ensure that all magnetization of the magnetic core is released to zero, so that the transformer will not saturate after multiple switching cycles. This energy released from the core can be achieved through several different technologies. A popular method is to use active clamps with small additional switches and capacitors.

Figure 7 shows a schematic diagram of the LTspice simulation environment using the ADP1074 forward active clamp design. In a forward converter, there is an additional inductor in the output path compared to a flyback converter, as shown in Figure 6. Although this is an additional component with related space and cost implications, it is not the same as a flyback converter. In addition, the transformer size required for a forward converter with the same power level as the flyback may be much smaller.

Figure 7: A positive active clamp circuit using the ADP1074 to generate an isolated output voltage, as simulated in LTspice.

In addition to flyback and forward topologies, there are many different transformer-based galvanically isolated converter concepts. The following list gives some very basic explanations about the most common converters:

The push-pull topology is similar to the forward converter. However, this topology does not require one low-side switch, but two active low-side switches. In addition, it requires a primary transformer winding with a center tap. The advantage of push-pull is that it generally has lower noise operation compared to forward converters and also requires a smaller transformer. The hysteresis of the transformer BH curve acts in two quadrants instead of one.

These two topologies are usually used for higher power designs ranging from a few hundred watts to several kilowatts. In addition to low-side switches, they also require high-side switches, but relatively small transformers can be used to achieve very high power transmission.

This term often appears when discussing high-power isolated converters. It stands for zero voltage switching. Another term for this type of converter is LLC (Inductor-Inductor-Capacitor) converter. These architectures are designed to achieve very efficient conversion. They generate a resonant circuit and switch the power switch when the voltage or current across the switch is close to zero. Therefore, switching losses are minimized. However, this type of design may be difficult to design and the switching frequency is not fixed, sometimes causing EMI problems.

In addition to linear regulators and switching power supplies, there is a third group of power converters: switched capacitor converters. They are also called charge pumps. They use switches and capacitors to multiply or reverse the voltage. They provide a huge advantage of not requiring any inductors. Generally, this type of converter is used for low power levels below 5 W. However, significant progress has recently been made, allowing the use of higher power switched capacitor converters. Figure 8 shows the LTC7820 designed with 120W, with an efficiency of 98.5%, converting 48V to 24V.

Figure 8: LTC7820 fixed ratio high power charge pump DC-DC controller.

All the power supplies discussed in this article can be implemented as analog or digital power supplies. What is a digital power supply? The power supply must always pass through an analog power stage with switches, inductors, transformers, and capacitors. The digital aspect is introduced by two digital building blocks. The first is a digital interface, which allows electronic systems to "talk" and "listen" to power. Different parameters can be set instantly to optimize the power supply for different operating conditions. In addition, the power supply can communicate with the main processor and issue warning or fault signs. For example, the system can easily monitor the load current, exceed a preset threshold, or the power supply temperature is too high.

The second digital building block replaces the analog regulation loop with a digital loop. This works successfully, but for most applications, the best option is a standard analog feedback loop, which has some digital effects on certain parameters, such as adjusting the gain of the error amplifier on the fly or dynamically setting loop compensation parameters to enable stable but fast Feedback loop. An example of a device with a purely digital control loop is the ADP1046A from Analog Devices. The LTC3883 is an example of a digital interface buck regulator with an analog control loop and optimized by digital influence.

When designing a switching power supply, electromagnetic interference (EMI) is always a topic that needs attention. The reason is that the switch mode power supply turns on and off large currents in a short period of time. The faster the switching speed, the higher the overall system efficiency. The faster switching transition reduces the time that the switch section is turned on. During this part of the on-time, most of the switching losses will occur. Figure 9 shows the switching node waveform of the switch mode power supply. Let us imagine a buck regulator. High voltage is defined by the current flowing through the high-side switch, and low voltage is defined by no current flowing through the high-side switch.

Figure 9: Switching conversion speed and switching frequency of a switch-mode power supply.

In Figure 9, we can see that the switch mode power supply will not only generate noise from the adjusted switching frequency, but also noise from higher frequency switching speeds. Although the switching frequency usually runs between 500kHz and 3MHz, the switching transition time may be only a few nanoseconds long. At the switching time of 1 ns, we will see the frequency corresponding to 1 GHz in the spectrum. At least these two frequencies will be considered radiated and conducted emissions. Other frequencies may also come from the oscillation of the regulation loop or the interaction between the power supply and the filter.

There are two reasons for reducing EMI. The first reason is to protect the function of an electronic system powered by a specific power source. For example, the 16-bit ADC used in the system signal path should not pick up switching noise from the power supply. The second reason is to meet certain EMI regulations formulated by governments around the world to protect the reliable functions of different electronic systems at the same time.

There are two forms of EMI, radiated EMI and conducted EMI. The most effective way to reduce radiated EMI is to optimize the PCB layout and use technologies, such as ADI's Silent Switcher technology. Of course, it is also effective to put the circuit in a shielded metal box. However, this may be impractical and in most cases very expensive.

Conducted EMI is usually attenuated by additional filtering. Additional filtering used to reduce conducted emissions will be discussed in the next section.

The RC filter is a basic low-pass filter. However, in power supply design, each filter is nothing but an LC filter. Usually, just adding some series inductance is enough, because it will form an LC or CLC filter together with the input or output capacitor of the switch mode power supply. Sometimes only capacitors are used as filters. However, considering the parasitic inductance of power lines or traces, we also form an LC filter together with capacitors. The inductor L can be an inductor with a magnetic core or a ferrite bead. The purpose of the LC filter is actually a low-pass effect, allowing the DC power supply to pass through and attenuating higher frequency interference to a large extent. The LC filter has two poles, so we get a high frequency attenuation of 40 dB per decade. This filter has a relatively sharp drop. Designing filters is not rocket science; however, because the parasitic components of the circuit (such as trace inductance) can have an impact, modeling the filter also requires modeling the main parasitic effects. This can make the simulation filter very time consuming. Many designers with filter design experience know which filters are effective in the past, and they may iteratively optimize a filter for a new design.

In all filter design, not only need to consider small signal behavior, such as the transfer function of the filter in the Bode diagram, but also need to understand the effects of large signals. In any LC filter, power flows through the inductor. If the power is no longer needed at the output, the energy stored in the inductor needs to be transferred somewhere due to sudden load transients. It charges the capacitor of the filter. If the filter is not designed for this worst-case scenario, the stored power may cause voltage overshoot, which may damage the circuit.

Figure 10: Use LTpowerCAD to design an input filter for a buck regulator.

Finally, the filter has a certain impedance. This impedance interacts with the impedance of the power converter connected to the filter. This interaction may cause instability and oscillation. Simulation tools such as LTspice and LTpowerCAD from ADI can greatly help answer all these questions and design perfect filters. Figure 10 shows the graphical user interface of the filter designer in the LTpowerCAD design environment. With this tool, the filter design is very simple.

Radiation emission is difficult to stop. Need to use some metal materials for special shielding. This can be very expensive. For a long time, engineers have been looking for ways to reduce the radiation generated by switch-mode power supplies. A few years ago, Silent Switcher technology made a major breakthrough. By reducing the parasitic inductance in the thermal loop of the switch-mode power supply, and dividing the thermal loop into two and setting them in a very symmetrical manner, most of the radiated emissions will cancel each other out. Today, many Silent Switcher devices have much lower radiation emissions than traditional products. Reducing radiation can increase the switching speed without causing serious EMI losses. Faster switching can reduce switching losses, thereby achieving a higher switching frequency. An example of this innovation is the LTC3310S, which can operate at a switching frequency of 5 MHz, enabling an extremely compact design with extremely low-cost external components.

Figure 11: The LTC3310S Silent Switcher is designed for the lowest radiation.

Power management is necessary, but can also be enjoyable

In this tutorial, we have studied many aspects of power supply design, including different power supply topologies and their advantages and disadvantages. For power engineers, this information may be very basic, but for experts and non-experts, it is helpful to have software tools such as LTpowerCAD and LTspice to help with the design process. Using these tools, power converters can be designed and optimized in a short period of time. Hope this tutorial will inspire you to look forward to the next power supply design challenge.

Frederik Dostal studied microelectronics at the University of Erlangen, Germany. He has been in the power management business since 2001 and has been active in various application positions, including four years of work in Phoenix, Arizona, where he worked in switch mode power supplies. He joined ADI in 2009 as a field application engineer for power management at ADI in Munich. You can contact him at frederik.dostal@analog.com.

Win a folding backpack! Learn about remote SSD management in the post-pandemic era!

You must register or log in to post a comment.